├── .DS_Store ├── .github └── workflows │ ├── build.yml │ └── linux-build.yml ├── .gitignore ├── Data ├── FashionMNIST_labels.npz ├── MNIST_labels.npz ├── WEBKB_labels.npz ├── cifar10_labels.npz ├── cifar_labels.npz ├── karate.pkl ├── signMNIST_labels.npz ├── signmnist_labels.npz └── yalefaces_labels.npz ├── LICENSE ├── LabelPermutations ├── FashionMNISTModRate_permutations.npz ├── FashionMNISTMult1_permutations.npz ├── FashionMNISTMult2_permutations.npz ├── FashionMNISTMult3_permutations.npz ├── FashionMNISTMult4_permutations.npz ├── FashionMNISTMult5_permutations.npz ├── FashionMNIST_permutations.npz ├── MNISTModRate_permutations.npz ├── MNISTMult1_permutations.npz ├── MNISTMult2_permutations.npz ├── MNISTMult3_permutations.npz ├── MNISTMult4_permutations.npz ├── MNISTMult5_permutations.npz ├── MNIST_permutations.npz ├── MNISTnew_permutations.npz ├── WEBKB_permutations.npz ├── cifar10_permutations.npz ├── cifarModRate_permutations.npz ├── cifarMult1_permutations.npz ├── cifarMult2_permutations.npz ├── cifarMult3_permutations.npz ├── cifarMult4_permutations.npz ├── cifarMult5_permutations.npz └── cifar_permutations.npz ├── MANIFEST.in ├── README.md ├── ResultsFromPaper ├── FashionMNISTModRate_vae_k10_centeredkernel_accuracy.csv ├── FashionMNISTModRate_vae_k10_laplace_accuracy.csv ├── FashionMNISTModRate_vae_k10_mbo_accuracy.csv ├── FashionMNISTModRate_vae_k10_modularitymbo_accuracy.csv ├── FashionMNISTModRate_vae_k10_nearestneighbor_accuracy.csv ├── FashionMNISTModRate_vae_k10_p10laplace_none_accuracy.csv ├── FashionMNISTModRate_vae_k10_poisson_accuracy.csv ├── FashionMNISTModRate_vae_k10_poissonmbo_accuracy.csv ├── FashionMNISTModRate_vae_k10_randomwalk_accuracy.csv ├── FashionMNISTModRate_vae_k10_sparselabelpropagation_accuracy.csv ├── FashionMNISTModRate_vae_k10_volumembo_T0.100_V1.000_accuracy.csv ├── FashionMNISTModRate_vae_k10_wnll_accuracy.csv ├── FashionMNISTMult1_vae_k10_modularitymbo_accuracy.csv ├── FashionMNISTMult1_vae_k10_poisson_accuracy.csv ├── FashionMNISTMult2_vae_k10_modularitymbo_accuracy.csv ├── FashionMNISTMult2_vae_k10_poisson_accuracy.csv ├── FashionMNISTMult3_vae_k10_modularitymbo_accuracy.csv ├── FashionMNISTMult3_vae_k10_poisson_accuracy.csv ├── FashionMNISTMult4_vae_k10_modularitymbo_accuracy.csv ├── FashionMNISTMult4_vae_k10_poisson_accuracy.csv ├── FashionMNISTMult5_vae_k10_modularitymbo_accuracy.csv ├── FashionMNISTMult5_vae_k10_poisson_accuracy.csv ├── FashionMNIST_vae_k10_centeredkernel_accuracy.csv ├── FashionMNIST_vae_k10_laplace_accuracy.csv ├── FashionMNIST_vae_k10_mbo_accuracy.csv ├── FashionMNIST_vae_k10_modularitymbo_accuracy.csv ├── FashionMNIST_vae_k10_nearestneighbor_accuracy.csv ├── FashionMNIST_vae_k10_p10laplace_none_accuracy.csv ├── FashionMNIST_vae_k10_poisson_accuracy.csv ├── FashionMNIST_vae_k10_poisson_accuracy50trials.csv ├── FashionMNIST_vae_k10_poissonmbo_accuracy.csv ├── FashionMNIST_vae_k10_poissonvolume_accuracy.csv ├── FashionMNIST_vae_k10_randomwalk_accuracy.csv ├── FashionMNIST_vae_k10_sparselabelpropagation_accuracy.csv ├── FashionMNIST_vae_k10_volumembo_T0.100_V1.000_accuracy.csv ├── FashionMNIST_vae_k10_wnll_accuracy.csv ├── FashionMNIST_vae_k12_poisson_accuracy50trials.csv ├── FashionMNIST_vae_k14_poisson_accuracy50trials.csv ├── FashionMNIST_vae_k16_poisson_accuracy50trials.csv ├── FashionMNIST_vae_k18_poisson_accuracy50trials.csv ├── FashionMNIST_vae_k20_poisson_accuracy50trials.csv ├── FashionMNIST_vae_k6_poisson_accuracy50trials.csv ├── FashionMNIST_vae_k8_poisson_accuracy50trials.csv ├── MNISTModRate_vae_k10_centeredkernel_accuracy.csv ├── MNISTModRate_vae_k10_laplace_accuracy.csv ├── MNISTModRate_vae_k10_mbo_accuracy.csv ├── MNISTModRate_vae_k10_modularitymbo_accuracy.csv ├── MNISTModRate_vae_k10_nearestneighbor_accuracy.csv ├── MNISTModRate_vae_k10_p10laplace_none_accuracy.csv ├── MNISTModRate_vae_k10_poisson_accuracy.csv ├── MNISTModRate_vae_k10_poissonmbo_accuracy.csv ├── MNISTModRate_vae_k10_randomwalk_accuracy.csv ├── MNISTModRate_vae_k10_sparselabelpropagation_accuracy.csv ├── MNISTModRate_vae_k10_volumembo_T0.100_V1.000_accuracy.csv ├── MNISTModRate_vae_k10_wnll_accuracy.csv ├── MNISTMult1_vae_k10_modularitymbo_accuracy.csv ├── MNISTMult1_vae_k10_poisson_accuracy.csv ├── MNISTMult2_vae_k10_modularitymbo_accuracy.csv ├── MNISTMult2_vae_k10_poisson_accuracy.csv ├── MNISTMult3_vae_k10_modularitymbo_accuracy.csv ├── MNISTMult3_vae_k10_poisson_accuracy.csv ├── MNISTMult4_vae_k10_modularitymbo_accuracy.csv ├── MNISTMult4_vae_k10_poisson_accuracy.csv ├── MNISTMult5_vae_k10_modularitymbo_accuracy.csv ├── MNISTMult5_vae_k10_poisson_accuracy.csv ├── MNIST_vae_k10_centeredkernel_accuracy.csv ├── MNIST_vae_k10_laplace_accuracy.csv ├── MNIST_vae_k10_mbo_accuracy.csv ├── MNIST_vae_k10_modularitymbo_accuracy.csv ├── MNIST_vae_k10_nearestneighbor_accuracy.csv ├── MNIST_vae_k10_p10laplace_none_accuracy.csv ├── MNIST_vae_k10_poisson_accuracy.csv ├── MNIST_vae_k10_poisson_accuracy50trials.csv ├── MNIST_vae_k10_poissonmbo_accuracy.csv ├── MNIST_vae_k10_poissonvolume_accuracy.csv ├── MNIST_vae_k10_randomwalk_accuracy.csv ├── MNIST_vae_k10_sparselabelpropagation_accuracy.csv ├── MNIST_vae_k10_volumembo_T0.100_V1.000_accuracy.csv ├── MNIST_vae_k10_wnll_accuracy.csv ├── MNIST_vae_k12_poisson_accuracy50trials.csv ├── MNIST_vae_k14_poisson_accuracy50trials.csv ├── MNIST_vae_k16_poisson_accuracy50trials.csv ├── MNIST_vae_k18_poisson_accuracy50trials.csv ├── MNIST_vae_k20_poisson_accuracy50trials.csv ├── MNIST_vae_k6_poisson_accuracy50trials.csv ├── MNIST_vae_k8_poisson_accuracy50trials.csv ├── cifarModRate_aet_k10_centeredkernel_accuracy.csv ├── cifarModRate_aet_k10_laplace_accuracy.csv ├── cifarModRate_aet_k10_mbo_accuracy.csv ├── cifarModRate_aet_k10_modularitymbo_accuracy.csv ├── cifarModRate_aet_k10_nearestneighbor_accuracy.csv ├── cifarModRate_aet_k10_p10laplace_none_accuracy.csv ├── cifarModRate_aet_k10_poisson_accuracy.csv ├── cifarModRate_aet_k10_poissonmbo_accuracy.csv ├── cifarModRate_aet_k10_randomwalk_accuracy.csv ├── cifarModRate_aet_k10_sparselabelpropagation_accuracy.csv ├── cifarModRate_aet_k10_volumembo_T0.100_V1.000_accuracy.csv ├── cifarModRate_aet_k10_wnll_accuracy.csv ├── cifarMult1_aet_k10_modularitymbo_accuracy.csv ├── cifarMult1_aet_k10_poisson_accuracy.csv ├── cifarMult2_aet_k10_modularitymbo_accuracy.csv ├── cifarMult2_aet_k10_poisson_accuracy.csv ├── cifarMult3_aet_k10_modularitymbo_accuracy.csv ├── cifarMult3_aet_k10_poisson_accuracy.csv ├── cifarMult4_aet_k10_modularitymbo_accuracy.csv ├── cifarMult4_aet_k10_poisson_accuracy.csv ├── cifarMult5_aet_k10_modularitymbo_accuracy.csv ├── cifarMult5_aet_k10_poisson_accuracy.csv ├── cifar_aet_k10_centeredkernel_accuracy.csv ├── cifar_aet_k10_laplace_accuracy.csv ├── cifar_aet_k10_mbo_accuracy.csv ├── cifar_aet_k10_modularitymbo_accuracy.csv ├── cifar_aet_k10_nearestneighbor_accuracy.csv ├── cifar_aet_k10_p10laplace_none_accuracy.csv ├── cifar_aet_k10_poisson_accuracy.csv ├── cifar_aet_k10_poisson_accuracy50trials.csv ├── cifar_aet_k10_poissonmbo_accuracy.csv ├── cifar_aet_k10_randomwalk_accuracy.csv ├── cifar_aet_k10_sparselabelpropagation_accuracy.csv ├── cifar_aet_k10_volumembo_T0.100_V1.000_accuracy.csv ├── cifar_aet_k10_wnll_accuracy.csv ├── cifar_aet_k12_poisson_accuracy50trials.csv ├── cifar_aet_k14_poisson_accuracy50trials.csv ├── cifar_aet_k16_poisson_accuracy50trials.csv ├── cifar_aet_k18_poisson_accuracy50trials.csv ├── cifar_aet_k20_poisson_accuracy50trials.csv ├── cifar_aet_k6_poisson_accuracy50trials.csv └── cifar_aet_k8_poisson_accuracy50trials.csv ├── c_code ├── LICENSE.txt ├── cextensions.cpp ├── hjsolvers.cpp ├── hjsolvers.h ├── lp_iterate.cpp ├── lp_iterate.h ├── maj_dijkstra.h ├── maj_implicit_heap.h ├── maj_simple_implicit_heap.h ├── mbo_convolution.cpp ├── mbo_convolution.h ├── memory_allocation.cpp ├── memory_allocation.h ├── mnist_benchmark.cpp ├── mnist_benchmark.h ├── sptree.cpp ├── sptree.h ├── tsne.cpp ├── tsne.h ├── vector_operations.h └── vptree.h ├── docs ├── _config.yml ├── active_learning.html ├── clustering.html ├── datasets.html ├── graph.html ├── index.html ├── ssl.html ├── trainsets.html ├── utils.html └── weightmatrix.html ├── examples ├── RP1D_mnist.py ├── al_test.py ├── ars_tsne.py ├── cone.png ├── dijkstra.py ├── dijkstra_hl.py ├── fokker_planck_clustering.py ├── incres_mnist.py ├── peikonal.py ├── plaplace.py ├── poisson_directed.py ├── poisson_mbo.py ├── randomized_svd.py ├── regression.py ├── spectral_mnist.py ├── spectral_twomoons.py ├── ssl_classpriors.py ├── ssl_mnist.py ├── ssl_trials.py ├── ssl_twomoons.py └── vae_mnist.py ├── generate_docs.sh ├── graphlearning ├── __init__.py ├── active_learning.py ├── clustering.py ├── datasets.py ├── graph.py ├── ssl.py ├── trainsets.py ├── utils.py └── weightmatrix.py ├── images ├── cluster_blobs.png ├── cluster_circles.png ├── cluster_moons.png └── clustering.png ├── kNNData ├── FashionMNIST_raw.npz ├── FashionMNIST_scatter.npz ├── FashionMNIST_vae.npz ├── MNIST_raw.npz ├── MNIST_scatter.npz ├── MNIST_vae.npz ├── MNIST_vae_old.npz ├── WEBKB_raw.npz ├── cifar10_simclr.npz ├── cifar10_simclr18.npz ├── cifar_aet.npz └── cifar_raw.npz ├── makefile ├── pdoc └── templates │ ├── _lunr_search.inc.mako │ ├── config.mako │ ├── credits.mako │ ├── css.mako │ ├── head.mako │ ├── html.mako │ ├── logo.mako │ ├── pdf.mako │ ├── search.mako │ └── text.mako ├── pyproject.toml ├── requirements.txt └── setup.py /.DS_Store: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/.DS_Store -------------------------------------------------------------------------------- /.github/workflows/build.yml: -------------------------------------------------------------------------------- 1 | name: Build 2 | 3 | on: 4 | push: 5 | branches: [master] 6 | 7 | jobs: 8 | build: 9 | runs-on: ${{ matrix.os }} 10 | strategy: 11 | matrix: 12 | os: [macos-latest, windows-latest] 13 | python-version: ['3.8','3.9','3.10','3.11','3.12'] 14 | 15 | steps: 16 | - uses: actions/checkout@v4 17 | - name: Set up Python 18 | uses: actions/setup-python@v5 19 | with: 20 | python-version: ${{ matrix.python-version }} 21 | - name: Install dependencies 22 | run: | 23 | python -m pip install --upgrade pip 24 | pip install setuptools wheel twine numpy build 25 | - name: Build and publish 26 | env: 27 | TWINE_USERNAME: __token__ 28 | TWINE_PASSWORD: ${{ secrets.PYPI_PASSWORD }} 29 | run: | 30 | python -m build --sdist --wheel 31 | twine upload --skip-existing dist/* 32 | -------------------------------------------------------------------------------- /.github/workflows/linux-build.yml: -------------------------------------------------------------------------------- 1 | name: Linux Build 2 | 3 | on: 4 | push: 5 | branches: [master] 6 | 7 | jobs: 8 | deploy: 9 | runs-on: ubuntu-latest 10 | 11 | steps: 12 | - uses: actions/checkout@v4 13 | - name: Set up Python 14 | uses: actions/setup-python@v5 15 | with: 16 | python-version: '3.x' 17 | - name: Install dependencies 18 | run: | 19 | python -m pip install --upgrade pip 20 | pip install setuptools wheel twine numpy build 21 | - name: Build and publish 22 | env: 23 | TWINE_USERNAME: __token__ 24 | TWINE_PASSWORD: ${{ secrets.PYPI_PASSWORD }} 25 | run: | 26 | python -m build --sdist 27 | twine upload --skip-existing dist/* 28 | -------------------------------------------------------------------------------- /.gitignore: -------------------------------------------------------------------------------- 1 | dist/ 2 | build/ 3 | graphlearning.egg-info/ 4 | graphlearning/__pycache__/ 5 | -------------------------------------------------------------------------------- /Data/FashionMNIST_labels.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/Data/FashionMNIST_labels.npz -------------------------------------------------------------------------------- /Data/MNIST_labels.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/Data/MNIST_labels.npz -------------------------------------------------------------------------------- /Data/WEBKB_labels.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/Data/WEBKB_labels.npz -------------------------------------------------------------------------------- /Data/cifar10_labels.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/Data/cifar10_labels.npz -------------------------------------------------------------------------------- /Data/cifar_labels.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/Data/cifar_labels.npz -------------------------------------------------------------------------------- /Data/karate.pkl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/Data/karate.pkl -------------------------------------------------------------------------------- /Data/signMNIST_labels.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/Data/signMNIST_labels.npz -------------------------------------------------------------------------------- /Data/signmnist_labels.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/Data/signmnist_labels.npz -------------------------------------------------------------------------------- /Data/yalefaces_labels.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/Data/yalefaces_labels.npz -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- 1 | MIT License 2 | 3 | Copyright (c) 2020 jwcalder 4 | 5 | Permission is hereby granted, free of charge, to any person obtaining a copy 6 | of this software and associated documentation files (the "Software"), to deal 7 | in the Software without restriction, including without limitation the rights 8 | to use, copy, modify, merge, publish, distribute, sublicense, and/or sell 9 | copies of the Software, and to permit persons to whom the Software is 10 | furnished to do so, subject to the following conditions: 11 | 12 | The above copyright notice and this permission notice shall be included in all 13 | copies or substantial portions of the Software. 14 | 15 | THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR 16 | IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, 17 | FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE 18 | AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER 19 | LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, 20 | OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE 21 | SOFTWARE. 22 | -------------------------------------------------------------------------------- /LabelPermutations/FashionMNISTModRate_permutations.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/LabelPermutations/FashionMNISTModRate_permutations.npz -------------------------------------------------------------------------------- /LabelPermutations/FashionMNISTMult1_permutations.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/LabelPermutations/FashionMNISTMult1_permutations.npz -------------------------------------------------------------------------------- /LabelPermutations/FashionMNISTMult2_permutations.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/LabelPermutations/FashionMNISTMult2_permutations.npz -------------------------------------------------------------------------------- /LabelPermutations/FashionMNISTMult3_permutations.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/LabelPermutations/FashionMNISTMult3_permutations.npz -------------------------------------------------------------------------------- /LabelPermutations/FashionMNISTMult4_permutations.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/LabelPermutations/FashionMNISTMult4_permutations.npz -------------------------------------------------------------------------------- /LabelPermutations/FashionMNISTMult5_permutations.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/LabelPermutations/FashionMNISTMult5_permutations.npz -------------------------------------------------------------------------------- /LabelPermutations/FashionMNIST_permutations.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/LabelPermutations/FashionMNIST_permutations.npz -------------------------------------------------------------------------------- /LabelPermutations/MNISTModRate_permutations.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/LabelPermutations/MNISTModRate_permutations.npz -------------------------------------------------------------------------------- /LabelPermutations/MNISTMult1_permutations.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/LabelPermutations/MNISTMult1_permutations.npz -------------------------------------------------------------------------------- /LabelPermutations/MNISTMult2_permutations.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/LabelPermutations/MNISTMult2_permutations.npz -------------------------------------------------------------------------------- /LabelPermutations/MNISTMult3_permutations.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/LabelPermutations/MNISTMult3_permutations.npz -------------------------------------------------------------------------------- /LabelPermutations/MNISTMult4_permutations.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/LabelPermutations/MNISTMult4_permutations.npz -------------------------------------------------------------------------------- /LabelPermutations/MNISTMult5_permutations.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/LabelPermutations/MNISTMult5_permutations.npz -------------------------------------------------------------------------------- /LabelPermutations/MNIST_permutations.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/LabelPermutations/MNIST_permutations.npz -------------------------------------------------------------------------------- /LabelPermutations/MNISTnew_permutations.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/LabelPermutations/MNISTnew_permutations.npz -------------------------------------------------------------------------------- /LabelPermutations/WEBKB_permutations.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/LabelPermutations/WEBKB_permutations.npz -------------------------------------------------------------------------------- /LabelPermutations/cifar10_permutations.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/LabelPermutations/cifar10_permutations.npz -------------------------------------------------------------------------------- /LabelPermutations/cifarModRate_permutations.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/LabelPermutations/cifarModRate_permutations.npz -------------------------------------------------------------------------------- /LabelPermutations/cifarMult1_permutations.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/LabelPermutations/cifarMult1_permutations.npz -------------------------------------------------------------------------------- /LabelPermutations/cifarMult2_permutations.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/LabelPermutations/cifarMult2_permutations.npz -------------------------------------------------------------------------------- /LabelPermutations/cifarMult3_permutations.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/LabelPermutations/cifarMult3_permutations.npz -------------------------------------------------------------------------------- /LabelPermutations/cifarMult4_permutations.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/LabelPermutations/cifarMult4_permutations.npz -------------------------------------------------------------------------------- /LabelPermutations/cifarMult5_permutations.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/LabelPermutations/cifarMult5_permutations.npz -------------------------------------------------------------------------------- /LabelPermutations/cifar_permutations.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/LabelPermutations/cifar_permutations.npz -------------------------------------------------------------------------------- /MANIFEST.in: -------------------------------------------------------------------------------- 1 | graft c_code 2 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | ## Graph-based Clustering and Semi-Supervised Learning 2 | 3 | ![Clustering](https://github.com/jwcalder/GraphLearning/raw/master/images/clustering.png) 4 | 5 | This python package is devoted to efficient implementations of modern graph-based learning algorithms for semi-supervised learning, active learning, and clustering. The package implements many popular datasets (currently MNIST, FashionMNIST, and CIFAR-10) in a way that makes it simple for users to test out new algorithms and rapidly compare against existing methods. Full [documentation](https://jwcalder.github.io/GraphLearning/) is available, including detailed example scripts. 6 | 7 | This package also reproduces experiments from the paper 8 | 9 | J. Calder, B. Cook, M. Thorpe, D. Slepcev. [Poisson Learning: Graph Based Semi-Supervised Learning at Very Low Label Rates.](http://proceedings.mlr.press/v119/calder20a.html), Proceedings of the 37th International Conference on Machine Learning, PMLR 119:1306-1316, 2020. 10 | 11 | **Important:** The variational autoencoder embedding (VAE) has been improved since our ICML paper was published. If you wish to load the VAE graph or data from that paper use `metric=vae_old` instead of `metric=vae`. 12 | 13 | ## Installation 14 | 15 | Install with 16 | ```sh 17 | pip install graphlearning 18 | ``` 19 | Required packages will be installed automatically, and include numpy, scipy, sklearn, and matplotlib. Some features in the package rely on other packages, including [annoy](https://github.com/spotify/annoy) for approximate nearest neighbor searches, and [torch](https://github.com/pytorch/pytorch) for GPU acceleration. You will have to install these manually, if needed, with 20 | ```sh 21 | pip install annoy torch 22 | ``` 23 | It can be difficult to install annoy, depending on your operating system. 24 | 25 | To install the most recent version of GraphLearning from the github source, which is updated more frequently, run 26 | ```sh 27 | git clone https://github.com/jwcalder/GraphLearning 28 | cd GraphLearning 29 | pip install . 30 | ``` 31 | If you prefer to use ssh swap the first line with 32 | ```sh 33 | git clone git@github.com:jwcalder/GraphLearning.git 34 | ``` 35 | 36 | ## Documentation and Examples 37 | 38 | Full documentation for the package is available [here](https://jwcalder.github.io/GraphLearning/). The documentation includes examples of how to use the package. All example scripts linked from the documentation can be found in the examples folder. 39 | 40 | ## Older versions of GraphLearning 41 | 42 | This repository hosts the current version of the package, which is numbered >=1.0.0. This version is not backwards compatible with earlier versions of the package. The old version is archived [here](https://github.com/jwcalder/GraphLearningOld) and can be installed with 43 | ```sh 44 | pip install graphlearning==0.0.3 45 | ``` 46 | To make sure you will load the old version when running `import graphlearning`, it may be necessary to uninstall all existing versions `pip uninstall graphlearning` before running the installation command above. 47 | 48 | ## Citations 49 | 50 | If you use this package in your research, please cite the package with the bibtex entry below. 51 | ``` 52 | @software{graphlearning, 53 | author = {Jeff Calder}, 54 | title = {GraphLearning Python Package}, 55 | month = jan, 56 | year = 2022, 57 | publisher = {Zenodo}, 58 | doi = {10.5281/zenodo.5850940}, 59 | url = {https://doi.org/10.5281/zenodo.5850940} 60 | } 61 | ``` 62 | 63 | ## Contact and questions 64 | 65 | Email with any questions or comments. 66 | 67 | ## Acknowledgments 68 | 69 | Several people have contributed to the development of this software: 70 | 71 | 1. Mauricio Rios Flores (Machine Learning Researcher, Amazon) 72 | 2. Brendan Cook (PhD Candidate in Mathematics, University of Minnesota) 73 | 3. Matt Jacobs (Postdoc, UCLA) 74 | 4. Mahmood Ettehad (Postdoc, IMA) 75 | 5. Jason Setiadi 76 | 6. Kevin Miller (Postdoc, Oden Institute) 77 | 78 | -------------------------------------------------------------------------------- /ResultsFromPaper/FashionMNISTModRate_vae_k10_p10laplace_none_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-03_14:58 2 | 1600,80.56 3 | 800,79.75 4 | 800,80.73 5 | 1600,81.04 6 | 400,78.14 7 | 400,78.10 8 | 200,76.83 9 | 200,77.37 10 | 1600,80.79 11 | 800,79.54 12 | 100,72.33 13 | 100,72.99 14 | 400,78.14 15 | 100,73.28 16 | 200,75.25 17 | 1600,80.45 18 | 800,78.96 19 | 400,78.12 20 | 100,74.05 21 | 200,76.24 22 | 1600,81.46 23 | 400,78.28 24 | 1600,80.67 25 | 200,75.18 26 | 800,80.02 27 | 800,79.66 28 | 400,77.81 29 | 100,74.63 30 | 200,76.17 31 | 1600,80.96 32 | 800,79.51 33 | 400,78.17 34 | 200,75.60 35 | 100,72.79 36 | 100,73.24 37 | 400,78.04 38 | 800,79.35 39 | 1600,80.77 40 | 100,70.99 41 | 200,77.46 42 | 1600,81.15 43 | 800,79.17 44 | 400,77.30 45 | 1600,80.67 46 | 200,76.83 47 | 800,80.09 48 | 100,73.55 49 | 400,78.19 50 | 200,75.24 51 | 100,72.55 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/FashionMNISTModRate_vae_k10_sparselabelpropagation_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-03_12:49 2 | 100,14.47 3 | 400,13.62 4 | 1600,54.09 5 | 200,19.86 6 | 1600,52.68 7 | 400,12.89 8 | 100,20.96 9 | 800,13.64 10 | 200,15.15 11 | 800,13.75 12 | 100,18.09 13 | 200,17.05 14 | 100,15.57 15 | 400,13.83 16 | 1600,50.70 17 | 800,11.87 18 | 200,12.49 19 | 800,15.86 20 | 1600,52.07 21 | 400,13.14 22 | 100,14.86 23 | 400,17.45 24 | 800,16.51 25 | 200,17.51 26 | 1600,51.21 27 | 100,14.74 28 | 200,15.50 29 | 1600,47.51 30 | 400,15.48 31 | 800,13.45 32 | 400,16.85 33 | 100,14.40 34 | 200,12.97 35 | 1600,53.62 36 | 100,11.13 37 | 800,12.27 38 | 1600,49.50 39 | 800,13.84 40 | 400,14.29 41 | 200,16.52 42 | 100,14.42 43 | 200,15.19 44 | 400,14.03 45 | 800,13.16 46 | 1600,53.37 47 | 100,13.09 48 | 200,16.46 49 | 400,13.88 50 | 800,13.40 51 | 1600,54.49 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/FashionMNISTMult1_vae_k10_modularitymbo_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2021-03-22_23:02 2 | 10,45.07 3 | 10,60.72 4 | 10,54.95 5 | 10,57.31 6 | 10,51.43 7 | 10,35.62 8 | 10,52.10 9 | 10,56.93 10 | 10,57.31 11 | 10,39.74 12 | 10,53.28 13 | 10,50.84 14 | 10,46.42 15 | 10,61.33 16 | 10,49.36 17 | 10,55.94 18 | 10,50.10 19 | 10,54.76 20 | 10,52.15 21 | 10,51.87 22 | 10,51.91 23 | 10,36.12 24 | 10,52.53 25 | 10,66.47 26 | 10,57.09 27 | 10,50.78 28 | 10,57.13 29 | 10,46.98 30 | 10,56.66 31 | 10,66.38 32 | 10,56.31 33 | 10,63.98 34 | 10,46.47 35 | 10,54.41 36 | 10,49.64 37 | 10,56.95 38 | 10,54.83 39 | 10,56.37 40 | 10,54.19 41 | 10,56.55 42 | 10,49.90 43 | 10,54.95 44 | 10,65.41 45 | 10,58.93 46 | 10,58.99 47 | 10,49.01 48 | 10,64.13 49 | 10,54.43 50 | 10,44.64 51 | 10,56.37 52 | 10,59.42 53 | 10,42.35 54 | 10,62.82 55 | 10,58.37 56 | 10,49.06 57 | 10,57.88 58 | 10,54.53 59 | 10,45.77 60 | 10,56.57 61 | 10,56.28 62 | 10,55.07 63 | 10,46.83 64 | 10,54.94 65 | 10,46.77 66 | 10,64.15 67 | 10,67.08 68 | 10,63.16 69 | 10,42.04 70 | 10,46.49 71 | 10,58.22 72 | 10,56.11 73 | 10,52.55 74 | 10,48.61 75 | 10,52.17 76 | 10,56.66 77 | 10,54.28 78 | 10,55.08 79 | 10,44.73 80 | 10,53.11 81 | 10,58.21 82 | 10,59.09 83 | 10,60.28 84 | 10,55.33 85 | 10,61.06 86 | 10,55.23 87 | 10,62.61 88 | 10,51.88 89 | 10,60.12 90 | 10,51.99 91 | 10,60.62 92 | 10,57.93 93 | 10,58.98 94 | 10,58.31 95 | 10,55.48 96 | 10,59.50 97 | 10,58.41 98 | 10,62.05 99 | 10,59.75 100 | 10,56.73 101 | 10,46.76 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/FashionMNISTMult1_vae_k10_poisson_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-08_23:06 2 | 10,59.31 3 | 10,67.37 4 | 10,61.11 5 | 10,61.04 6 | 10,59.99 7 | 10,57.84 8 | 10,67.11 9 | 10,51.12 10 | 10,62.87 11 | 10,50.97 12 | 10,51.63 13 | 10,63.59 14 | 10,58.15 15 | 10,61.52 16 | 10,60.19 17 | 10,56.76 18 | 10,58.99 19 | 10,63.63 20 | 10,67.55 21 | 10,63.10 22 | 10,61.77 23 | 10,50.89 24 | 10,47.06 25 | 10,62.62 26 | 10,60.84 27 | 10,64.10 28 | 10,62.87 29 | 10,62.52 30 | 10,61.25 31 | 10,62.20 32 | 10,56.08 33 | 10,58.70 34 | 10,67.10 35 | 10,61.52 36 | 10,61.05 37 | 10,56.75 38 | 10,59.52 39 | 10,64.33 40 | 10,63.16 41 | 10,51.91 42 | 10,61.07 43 | 10,56.47 44 | 10,62.74 45 | 10,63.16 46 | 10,61.33 47 | 10,49.88 48 | 10,58.91 49 | 10,61.55 50 | 10,57.93 51 | 10,59.74 52 | 10,63.31 53 | 10,65.62 54 | 10,54.91 55 | 10,61.33 56 | 10,57.81 57 | 10,69.46 58 | 10,58.23 59 | 10,59.75 60 | 10,52.77 61 | 10,69.78 62 | 10,57.58 63 | 10,50.19 64 | 10,59.54 65 | 10,65.04 66 | 10,60.98 67 | 10,70.27 68 | 10,65.79 69 | 10,55.63 70 | 10,59.58 71 | 10,52.76 72 | 10,60.28 73 | 10,58.92 74 | 10,62.49 75 | 10,64.00 76 | 10,66.63 77 | 10,56.05 78 | 10,67.96 79 | 10,59.58 80 | 10,66.67 81 | 10,58.48 82 | 10,58.04 83 | 10,59.99 84 | 10,67.90 85 | 10,56.05 86 | 10,58.41 87 | 10,63.91 88 | 10,62.42 89 | 10,64.38 90 | 10,62.31 91 | 10,65.90 92 | 10,60.62 93 | 10,67.13 94 | 10,56.46 95 | 10,65.37 96 | 10,56.96 97 | 10,68.85 98 | 10,58.37 99 | 10,66.43 100 | 10,61.43 101 | 10,48.89 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/FashionMNISTMult2_vae_k10_modularitymbo_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2021-03-22_23:03 2 | 15,65.94 3 | 15,58.76 4 | 15,55.66 5 | 15,53.93 6 | 15,61.95 7 | 15,62.46 8 | 15,60.97 9 | 15,68.54 10 | 15,60.13 11 | 15,62.18 12 | 15,59.92 13 | 15,63.99 14 | 15,57.15 15 | 15,51.53 16 | 15,53.63 17 | 15,61.49 18 | 15,57.20 19 | 15,50.24 20 | 15,53.30 21 | 15,67.11 22 | 15,55.20 23 | 15,51.24 24 | 15,59.00 25 | 15,62.11 26 | 15,67.74 27 | 15,55.55 28 | 15,59.81 29 | 15,60.52 30 | 15,57.46 31 | 15,56.22 32 | 15,59.27 33 | 15,62.72 34 | 15,68.38 35 | 15,61.45 36 | 15,46.16 37 | 15,55.96 38 | 15,51.51 39 | 15,59.27 40 | 15,60.61 41 | 15,60.53 42 | 15,51.65 43 | 15,58.47 44 | 15,57.69 45 | 15,49.66 46 | 15,63.81 47 | 15,60.31 48 | 15,65.05 49 | 15,58.93 50 | 15,62.77 51 | 15,59.48 52 | 15,61.87 53 | 15,67.32 54 | 15,61.29 55 | 15,65.13 56 | 15,63.83 57 | 15,54.97 58 | 15,73.20 59 | 15,60.00 60 | 15,59.44 61 | 15,59.93 62 | 15,56.41 63 | 15,45.07 64 | 15,62.93 65 | 15,57.73 66 | 15,59.81 67 | 15,60.79 68 | 15,67.15 69 | 15,63.74 70 | 15,61.91 71 | 15,61.30 72 | 15,59.29 73 | 15,64.22 74 | 15,65.87 75 | 15,64.60 76 | 15,54.67 77 | 15,51.26 78 | 15,54.08 79 | 15,67.72 80 | 15,60.91 81 | 15,58.32 82 | 15,63.67 83 | 15,62.53 84 | 15,62.79 85 | 15,61.75 86 | 15,54.26 87 | 15,61.40 88 | 15,56.30 89 | 15,61.45 90 | 15,53.73 91 | 15,60.28 92 | 15,59.13 93 | 15,67.79 94 | 15,66.20 95 | 15,66.92 96 | 15,61.60 97 | 15,66.43 98 | 15,52.93 99 | 15,57.94 100 | 15,59.98 101 | 15,49.47 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/FashionMNISTMult2_vae_k10_poisson_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-08_23:18 2 | 15,65.57 3 | 15,64.62 4 | 15,64.08 5 | 15,65.34 6 | 15,68.73 7 | 15,64.37 8 | 15,68.41 9 | 15,59.75 10 | 15,66.79 11 | 15,64.78 12 | 15,67.00 13 | 15,58.71 14 | 15,58.50 15 | 15,68.33 16 | 15,61.46 17 | 15,57.84 18 | 15,67.59 19 | 15,61.15 20 | 15,70.57 21 | 15,64.29 22 | 15,62.21 23 | 15,59.32 24 | 15,61.22 25 | 15,55.27 26 | 15,55.77 27 | 15,69.49 28 | 15,65.96 29 | 15,64.74 30 | 15,67.21 31 | 15,64.97 32 | 15,68.86 33 | 15,56.88 34 | 15,63.14 35 | 15,56.39 36 | 15,66.42 37 | 15,65.79 38 | 15,61.24 39 | 15,69.06 40 | 15,57.77 41 | 15,60.95 42 | 15,61.72 43 | 15,58.32 44 | 15,64.97 45 | 15,70.09 46 | 15,68.44 47 | 15,66.10 48 | 15,54.98 49 | 15,68.17 50 | 15,59.46 51 | 15,54.64 52 | 15,64.91 53 | 15,66.92 54 | 15,63.03 55 | 15,62.38 56 | 15,68.24 57 | 15,63.79 58 | 15,69.23 59 | 15,63.16 60 | 15,59.09 61 | 15,66.97 62 | 15,57.30 63 | 15,63.90 64 | 15,52.32 65 | 15,56.10 66 | 15,63.27 67 | 15,64.79 68 | 15,66.14 69 | 15,62.81 70 | 15,65.05 71 | 15,65.24 72 | 15,54.68 73 | 15,67.15 74 | 15,60.84 75 | 15,61.36 76 | 15,68.34 77 | 15,60.23 78 | 15,65.52 79 | 15,65.05 80 | 15,66.54 81 | 15,71.95 82 | 15,60.31 83 | 15,62.50 84 | 15,70.62 85 | 15,67.18 86 | 15,62.97 87 | 15,59.59 88 | 15,70.05 89 | 15,61.11 90 | 15,62.09 91 | 15,72.02 92 | 15,68.34 93 | 15,63.78 94 | 15,64.68 95 | 15,68.16 96 | 15,66.96 97 | 15,68.33 98 | 15,57.18 99 | 15,71.54 100 | 15,60.29 101 | 15,62.98 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/FashionMNISTMult3_vae_k10_modularitymbo_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2021-03-22_23:03 2 | 20,62.67 3 | 20,67.39 4 | 20,50.21 5 | 20,50.73 6 | 20,66.02 7 | 20,63.40 8 | 20,63.93 9 | 20,59.28 10 | 20,54.95 11 | 20,66.65 12 | 20,66.58 13 | 20,50.40 14 | 20,56.02 15 | 20,54.95 16 | 20,56.06 17 | 20,58.99 18 | 20,59.02 19 | 20,60.09 20 | 20,56.71 21 | 20,57.99 22 | 20,63.81 23 | 20,66.75 24 | 20,49.27 25 | 20,59.13 26 | 20,68.92 27 | 20,62.80 28 | 20,64.97 29 | 20,56.95 30 | 20,64.05 31 | 20,65.00 32 | 20,59.99 33 | 20,69.45 34 | 20,62.61 35 | 20,65.74 36 | 20,57.85 37 | 20,58.53 38 | 20,61.54 39 | 20,66.10 40 | 20,65.87 41 | 20,65.01 42 | 20,60.57 43 | 20,57.18 44 | 20,65.41 45 | 20,65.09 46 | 20,55.36 47 | 20,62.76 48 | 20,63.52 49 | 20,56.14 50 | 20,64.89 51 | 20,62.84 52 | 20,55.04 53 | 20,58.61 54 | 20,62.70 55 | 20,73.22 56 | 20,58.76 57 | 20,56.97 58 | 20,55.37 59 | 20,62.08 60 | 20,63.37 61 | 20,58.69 62 | 20,56.82 63 | 20,51.74 64 | 20,65.09 65 | 20,65.32 66 | 20,63.09 67 | 20,67.71 68 | 20,57.34 69 | 20,69.04 70 | 20,56.67 71 | 20,57.37 72 | 20,54.90 73 | 20,68.79 74 | 20,53.94 75 | 20,59.26 76 | 20,58.04 77 | 20,61.89 78 | 20,64.02 79 | 20,68.45 80 | 20,67.47 81 | 20,53.74 82 | 20,59.60 83 | 20,66.08 84 | 20,67.37 85 | 20,59.58 86 | 20,67.60 87 | 20,61.73 88 | 20,62.22 89 | 20,63.57 90 | 20,65.81 91 | 20,55.70 92 | 20,70.64 93 | 20,64.24 94 | 20,59.88 95 | 20,56.54 96 | 20,63.86 97 | 20,62.26 98 | 20,62.40 99 | 20,61.27 100 | 20,63.52 101 | 20,69.84 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/FashionMNISTMult3_vae_k10_poisson_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-08_23:33 2 | 20,68.27 3 | 20,63.43 4 | 20,64.83 5 | 20,69.95 6 | 20,66.47 7 | 20,68.59 8 | 20,64.04 9 | 20,69.38 10 | 20,68.81 11 | 20,58.63 12 | 20,65.19 13 | 20,64.05 14 | 20,57.76 15 | 20,66.34 16 | 20,59.86 17 | 20,61.56 18 | 20,59.72 19 | 20,61.21 20 | 20,67.09 21 | 20,59.55 22 | 20,67.21 23 | 20,69.24 24 | 20,67.86 25 | 20,62.01 26 | 20,71.37 27 | 20,66.56 28 | 20,66.43 29 | 20,65.95 30 | 20,64.55 31 | 20,63.51 32 | 20,66.01 33 | 20,69.14 34 | 20,68.90 35 | 20,62.77 36 | 20,66.52 37 | 20,70.96 38 | 20,70.17 39 | 20,66.29 40 | 20,64.84 41 | 20,65.22 42 | 20,66.89 43 | 20,61.53 44 | 20,67.73 45 | 20,68.37 46 | 20,66.55 47 | 20,71.27 48 | 20,57.03 49 | 20,63.37 50 | 20,69.88 51 | 20,54.21 52 | 20,65.38 53 | 20,63.05 54 | 20,63.38 55 | 20,64.45 56 | 20,60.34 57 | 20,71.19 58 | 20,66.13 59 | 20,57.57 60 | 20,64.73 61 | 20,61.25 62 | 20,68.02 63 | 20,62.88 64 | 20,68.91 65 | 20,54.57 66 | 20,63.82 67 | 20,70.66 68 | 20,64.41 69 | 20,69.06 70 | 20,62.05 71 | 20,64.51 72 | 20,66.80 73 | 20,61.14 74 | 20,66.34 75 | 20,63.89 76 | 20,63.91 77 | 20,63.99 78 | 20,66.71 79 | 20,68.25 80 | 20,68.49 81 | 20,67.84 82 | 20,61.06 83 | 20,65.16 84 | 20,60.57 85 | 20,62.22 86 | 20,68.65 87 | 20,63.71 88 | 20,71.37 89 | 20,69.16 90 | 20,68.29 91 | 20,62.86 92 | 20,65.25 93 | 20,69.91 94 | 20,68.18 95 | 20,63.12 96 | 20,61.83 97 | 20,62.00 98 | 20,69.78 99 | 20,63.82 100 | 20,68.75 101 | 20,69.37 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/FashionMNISTMult4_vae_k10_modularitymbo_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2021-03-22_23:04 2 | 25,65.70 3 | 25,54.77 4 | 25,62.97 5 | 25,68.27 6 | 25,68.97 7 | 25,62.34 8 | 25,70.29 9 | 25,69.89 10 | 25,57.39 11 | 25,64.25 12 | 25,70.11 13 | 25,56.94 14 | 25,65.70 15 | 25,57.74 16 | 25,65.63 17 | 25,63.76 18 | 25,65.85 19 | 25,66.33 20 | 25,68.06 21 | 25,47.31 22 | 25,63.30 23 | 25,61.17 24 | 25,66.74 25 | 25,60.66 26 | 25,72.46 27 | 25,66.80 28 | 25,69.09 29 | 25,50.45 30 | 25,60.31 31 | 25,62.81 32 | 25,64.79 33 | 25,56.73 34 | 25,62.21 35 | 25,67.44 36 | 25,63.09 37 | 25,60.70 38 | 25,63.05 39 | 25,55.51 40 | 25,70.71 41 | 25,70.64 42 | 25,64.98 43 | 25,57.59 44 | 25,66.39 45 | 25,70.76 46 | 25,51.47 47 | 25,65.45 48 | 25,64.73 49 | 25,51.93 50 | 25,63.49 51 | 25,58.56 52 | 25,62.18 53 | 25,62.17 54 | 25,61.96 55 | 25,61.01 56 | 25,60.50 57 | 25,55.96 58 | 25,69.43 59 | 25,57.72 60 | 25,60.73 61 | 25,54.73 62 | 25,67.57 63 | 25,65.90 64 | 25,61.49 65 | 25,57.54 66 | 25,65.25 67 | 25,68.65 68 | 25,65.44 69 | 25,71.11 70 | 25,63.86 71 | 25,69.05 72 | 25,61.89 73 | 25,54.07 74 | 25,59.17 75 | 25,68.89 76 | 25,59.79 77 | 25,62.46 78 | 25,63.04 79 | 25,63.50 80 | 25,63.59 81 | 25,51.87 82 | 25,54.45 83 | 25,59.06 84 | 25,62.71 85 | 25,65.57 86 | 25,65.88 87 | 25,57.06 88 | 25,54.86 89 | 25,57.23 90 | 25,57.11 91 | 25,57.07 92 | 25,58.15 93 | 25,60.85 94 | 25,66.91 95 | 25,56.13 96 | 25,58.75 97 | 25,63.87 98 | 25,53.85 99 | 25,67.44 100 | 25,67.26 101 | 25,63.27 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/FashionMNISTMult4_vae_k10_poisson_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-08_23:52 2 | 25,62.91 3 | 25,65.82 4 | 25,66.37 5 | 25,66.15 6 | 25,70.34 7 | 25,68.30 8 | 25,66.97 9 | 25,60.49 10 | 25,71.66 11 | 25,63.93 12 | 25,67.46 13 | 25,68.88 14 | 25,65.52 15 | 25,70.31 16 | 25,63.81 17 | 25,68.14 18 | 25,71.68 19 | 25,66.15 20 | 25,59.86 21 | 25,60.62 22 | 25,72.51 23 | 25,62.27 24 | 25,64.94 25 | 25,69.37 26 | 25,66.37 27 | 25,70.46 28 | 25,68.62 29 | 25,57.94 30 | 25,59.68 31 | 25,65.88 32 | 25,67.81 33 | 25,62.69 34 | 25,68.57 35 | 25,65.97 36 | 25,67.44 37 | 25,62.06 38 | 25,65.80 39 | 25,73.41 40 | 25,63.66 41 | 25,65.28 42 | 25,64.25 43 | 25,60.43 44 | 25,67.51 45 | 25,69.56 46 | 25,69.96 47 | 25,59.76 48 | 25,54.90 49 | 25,65.94 50 | 25,60.54 51 | 25,67.45 52 | 25,68.13 53 | 25,61.97 54 | 25,63.10 55 | 25,60.73 56 | 25,58.94 57 | 25,70.84 58 | 25,61.77 59 | 25,63.36 60 | 25,67.46 61 | 25,64.35 62 | 25,69.50 63 | 25,69.06 64 | 25,61.09 65 | 25,69.18 66 | 25,69.65 67 | 25,62.72 68 | 25,67.95 69 | 25,75.26 70 | 25,68.91 71 | 25,69.15 72 | 25,55.34 73 | 25,70.37 74 | 25,67.21 75 | 25,69.22 76 | 25,68.00 77 | 25,61.13 78 | 25,64.06 79 | 25,67.99 80 | 25,68.39 81 | 25,63.81 82 | 25,60.95 83 | 25,66.72 84 | 25,67.86 85 | 25,66.93 86 | 25,69.44 87 | 25,67.05 88 | 25,63.69 89 | 25,65.79 90 | 25,56.30 91 | 25,60.70 92 | 25,62.99 93 | 25,66.45 94 | 25,64.93 95 | 25,71.01 96 | 25,59.61 97 | 25,56.73 98 | 25,68.79 99 | 25,63.67 100 | 25,55.69 101 | 25,66.10 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/FashionMNISTMult5_vae_k10_modularitymbo_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2021-03-22_23:04 2 | 30,64.43 3 | 30,68.35 4 | 30,54.89 5 | 30,62.02 6 | 30,70.52 7 | 30,68.03 8 | 30,71.68 9 | 30,62.79 10 | 30,60.04 11 | 30,64.05 12 | 30,65.62 13 | 30,62.59 14 | 30,64.47 15 | 30,64.21 16 | 30,64.68 17 | 30,65.17 18 | 30,66.71 19 | 30,64.55 20 | 30,64.72 21 | 30,67.55 22 | 30,56.03 23 | 30,69.41 24 | 30,50.57 25 | 30,64.76 26 | 30,70.14 27 | 30,56.49 28 | 30,70.30 29 | 30,62.52 30 | 30,64.15 31 | 30,62.87 32 | 30,60.10 33 | 30,67.48 34 | 30,65.40 35 | 30,65.13 36 | 30,61.72 37 | 30,69.97 38 | 30,63.18 39 | 30,62.27 40 | 30,59.96 41 | 30,57.75 42 | 30,70.97 43 | 30,70.85 44 | 30,61.14 45 | 30,67.23 46 | 30,62.24 47 | 30,67.10 48 | 30,64.06 49 | 30,59.33 50 | 30,60.16 51 | 30,63.77 52 | 30,65.28 53 | 30,65.17 54 | 30,66.63 55 | 30,68.54 56 | 30,63.08 57 | 30,66.17 58 | 30,68.49 59 | 30,65.52 60 | 30,67.37 61 | 30,59.36 62 | 30,65.42 63 | 30,59.97 64 | 30,65.78 65 | 30,67.71 66 | 30,70.58 67 | 30,70.41 68 | 30,63.89 69 | 30,64.31 70 | 30,59.87 71 | 30,66.29 72 | 30,61.80 73 | 30,62.14 74 | 30,63.30 75 | 30,67.10 76 | 30,64.78 77 | 30,65.39 78 | 30,65.66 79 | 30,68.28 80 | 30,67.85 81 | 30,67.42 82 | 30,65.50 83 | 30,62.75 84 | 30,57.31 85 | 30,53.97 86 | 30,59.10 87 | 30,68.15 88 | 30,65.69 89 | 30,67.29 90 | 30,63.80 91 | 30,61.75 92 | 30,65.93 93 | 30,67.00 94 | 30,55.65 95 | 30,66.94 96 | 30,66.90 97 | 30,68.10 98 | 30,60.19 99 | 30,55.06 100 | 30,68.00 101 | 30,65.12 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/FashionMNISTMult5_vae_k10_poisson_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-09_00:13 2 | 30,68.60 3 | 30,62.43 4 | 30,70.85 5 | 30,64.07 6 | 30,64.52 7 | 30,55.90 8 | 30,67.89 9 | 30,69.99 10 | 30,64.04 11 | 30,68.01 12 | 30,67.48 13 | 30,68.78 14 | 30,62.80 15 | 30,66.64 16 | 30,65.20 17 | 30,67.15 18 | 30,66.46 19 | 30,67.24 20 | 30,65.45 21 | 30,62.97 22 | 30,68.67 23 | 30,68.79 24 | 30,61.79 25 | 30,67.22 26 | 30,70.82 27 | 30,68.94 28 | 30,62.92 29 | 30,71.65 30 | 30,63.95 31 | 30,64.27 32 | 30,66.00 33 | 30,67.92 34 | 30,71.95 35 | 30,65.42 36 | 30,71.76 37 | 30,69.28 38 | 30,66.93 39 | 30,70.37 40 | 30,58.90 41 | 30,70.30 42 | 30,65.95 43 | 30,67.12 44 | 30,63.42 45 | 30,69.11 46 | 30,67.99 47 | 30,66.55 48 | 30,62.90 49 | 30,66.66 50 | 30,69.97 51 | 30,65.82 52 | 30,70.01 53 | 30,67.87 54 | 30,63.67 55 | 30,70.21 56 | 30,69.20 57 | 30,66.20 58 | 30,67.56 59 | 30,67.54 60 | 30,66.17 61 | 30,67.96 62 | 30,62.31 63 | 30,64.91 64 | 30,67.06 65 | 30,67.58 66 | 30,67.10 67 | 30,66.11 68 | 30,64.91 69 | 30,70.21 70 | 30,67.58 71 | 30,69.32 72 | 30,66.67 73 | 30,65.84 74 | 30,65.18 75 | 30,65.30 76 | 30,68.46 77 | 30,65.97 78 | 30,62.89 79 | 30,70.52 80 | 30,68.38 81 | 30,73.76 82 | 30,66.98 83 | 30,58.93 84 | 30,62.36 85 | 30,61.57 86 | 30,63.71 87 | 30,66.27 88 | 30,67.51 89 | 30,64.42 90 | 30,65.58 91 | 30,65.08 92 | 30,63.89 93 | 30,58.54 94 | 30,67.70 95 | 30,67.88 96 | 30,62.54 97 | 30,65.73 98 | 30,71.31 99 | 30,63.37 100 | 30,67.76 101 | 30,70.04 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/FashionMNIST_vae_k10_mbo_accuracy.csv: -------------------------------------------------------------------------------- 1 | 10,9.98 2 | 20,22.29 3 | 30,22.98 4 | 40,34.89 5 | 50,29.83 6 | 10,15.93 7 | 20,24.70 8 | 30,28.30 9 | 40,32.97 10 | 50,28.02 11 | 10,18.21 12 | 20,17.49 13 | 30,32.72 14 | 40,21.99 15 | 50,30.06 16 | 10,20.10 17 | 20,26.01 18 | 30,25.92 19 | 40,23.35 20 | 50,36.28 21 | 10,16.79 22 | 20,24.20 23 | 30,29.25 24 | 40,30.32 25 | 50,28.10 26 | 10,14.82 27 | 20,20.20 28 | 30,22.26 29 | 40,28.52 30 | 50,38.89 31 | 10,13.63 32 | 20,22.69 33 | 30,21.34 34 | 40,30.09 35 | 50,24.76 36 | 10,22.60 37 | 20,20.58 38 | 30,22.14 39 | 40,37.11 40 | 50,30.45 41 | 10,13.48 42 | 20,18.88 43 | 30,25.56 44 | 40,30.10 45 | 50,38.79 46 | 10,21.78 47 | 20,18.62 48 | 30,18.47 49 | 40,28.12 50 | 50,35.85 51 | 10,17.62 52 | 20,13.39 53 | 30,24.37 54 | 40,24.96 55 | 50,41.11 56 | 10,13.66 57 | 20,11.81 58 | 30,32.39 59 | 40,25.59 60 | 50,41.19 61 | 10,18.95 62 | 20,18.01 63 | 30,20.77 64 | 40,32.73 65 | 50,28.57 66 | 10,17.30 67 | 20,17.71 68 | 30,16.89 69 | 40,29.24 70 | 50,36.55 71 | 10,14.68 72 | 20,17.08 73 | 30,27.44 74 | 40,22.44 75 | 50,37.71 76 | 10,11.49 77 | 20,23.17 78 | 30,32.89 79 | 40,34.99 80 | 50,36.32 81 | 10,17.24 82 | 20,18.08 83 | 30,27.83 84 | 40,20.64 85 | 50,35.48 86 | 10,18.05 87 | 20,20.44 88 | 30,25.20 89 | 40,34.48 90 | 50,29.25 91 | 10,14.36 92 | 20,20.52 93 | 30,32.36 94 | 40,37.68 95 | 50,35.64 96 | 10,12.09 97 | 20,21.93 98 | 30,21.00 99 | 40,32.02 100 | 50,37.68 101 | 10,12.68 102 | 20,23.31 103 | 30,21.02 104 | 40,34.09 105 | 50,28.36 106 | 10,16.16 107 | 20,17.38 108 | 30,29.67 109 | 40,35.48 110 | 50,41.64 111 | 10,15.44 112 | 20,25.34 113 | 30,22.83 114 | 40,32.87 115 | 50,31.28 116 | 10,9.57 117 | 20,22.82 118 | 30,27.51 119 | 40,33.05 120 | 50,36.98 121 | 10,17.71 122 | 20,19.67 123 | 30,22.78 124 | 40,31.06 125 | 50,31.32 126 | 10,15.08 127 | 20,19.73 128 | 30,24.20 129 | 40,35.42 130 | 50,35.23 131 | 10,17.72 132 | 20,21.04 133 | 30,22.43 134 | 40,39.16 135 | 50,33.47 136 | 10,17.97 137 | 20,24.76 138 | 30,19.23 139 | 40,20.53 140 | 50,34.59 141 | 10,13.18 142 | 20,21.12 143 | 30,20.85 144 | 40,34.41 145 | 50,34.73 146 | 10,13.49 147 | 20,22.85 148 | 30,33.05 149 | 40,34.54 150 | 50,41.18 151 | 10,15.25 152 | 20,35.44 153 | 30,26.08 154 | 40,24.15 155 | 50,32.29 156 | 10,5.65 157 | 20,12.70 158 | 30,37.37 159 | 40,28.87 160 | 50,38.98 161 | 10,16.21 162 | 20,19.50 163 | 30,27.01 164 | 40,31.03 165 | 50,38.52 166 | 10,16.86 167 | 20,12.59 168 | 30,17.71 169 | 40,32.12 170 | 50,41.71 171 | 10,15.11 172 | 20,25.19 173 | 30,24.46 174 | 40,24.87 175 | 50,36.10 176 | 10,15.15 177 | 20,19.29 178 | 30,22.57 179 | 40,30.94 180 | 50,33.72 181 | 10,12.65 182 | 20,18.15 183 | 30,25.59 184 | 40,24.96 185 | 50,37.28 186 | 10,11.74 187 | 20,24.80 188 | 30,20.76 189 | 40,29.60 190 | 50,37.80 191 | 10,10.18 192 | 20,16.96 193 | 30,27.43 194 | 40,34.83 195 | 50,33.41 196 | 10,18.68 197 | 20,15.71 198 | 30,23.27 199 | 40,34.24 200 | 50,32.99 201 | 10,9.06 202 | 20,20.10 203 | 30,27.09 204 | 40,26.54 205 | 50,38.53 206 | 10,18.42 207 | 20,17.33 208 | 30,14.88 209 | 40,38.20 210 | 50,34.71 211 | 10,13.38 212 | 20,17.36 213 | 30,24.82 214 | 40,35.23 215 | 50,33.36 216 | 10,20.87 217 | 20,29.16 218 | 30,26.66 219 | 40,30.72 220 | 50,33.71 221 | 10,7.89 222 | 20,24.24 223 | 30,31.30 224 | 40,29.14 225 | 50,36.74 226 | 10,20.08 227 | 20,16.25 228 | 30,20.07 229 | 40,23.92 230 | 50,30.76 231 | 10,18.87 232 | 20,19.96 233 | 30,36.93 234 | 40,26.56 235 | 50,39.57 236 | 10,20.19 237 | 20,19.41 238 | 30,27.99 239 | 40,39.93 240 | 50,40.27 241 | 10,19.92 242 | 20,14.42 243 | 30,25.18 244 | 40,33.12 245 | 50,31.58 246 | 10,18.89 247 | 20,18.41 248 | 30,28.06 249 | 40,32.87 250 | 50,35.53 251 | 10,19.40 252 | 20,17.02 253 | 30,28.22 254 | 40,36.98 255 | 50,40.37 256 | 10,13.09 257 | 20,24.05 258 | 30,22.66 259 | 40,39.76 260 | 50,38.43 261 | 10,17.17 262 | 20,28.31 263 | 30,26.05 264 | 40,27.15 265 | 50,30.43 266 | 10,21.18 267 | 20,21.05 268 | 30,23.83 269 | 40,23.44 270 | 50,34.67 271 | 10,13.48 272 | 20,21.34 273 | 30,25.49 274 | 40,26.53 275 | 50,40.06 276 | 10,19.01 277 | 20,25.94 278 | 30,26.11 279 | 40,24.54 280 | 50,40.13 281 | 10,14.49 282 | 20,15.34 283 | 30,19.75 284 | 40,31.04 285 | 50,36.67 286 | 10,14.08 287 | 20,22.38 288 | 30,27.76 289 | 40,25.22 290 | 50,31.38 291 | 10,11.73 292 | 20,17.48 293 | 30,26.85 294 | 40,34.47 295 | 50,34.89 296 | 10,18.20 297 | 20,28.05 298 | 30,26.45 299 | 40,28.67 300 | 50,27.82 301 | 10,10.63 302 | 20,22.13 303 | 30,23.59 304 | 40,26.97 305 | 50,31.61 306 | 10,16.92 307 | 20,31.20 308 | 30,20.14 309 | 40,22.61 310 | 50,26.28 311 | 10,12.28 312 | 20,15.43 313 | 30,37.90 314 | 40,28.25 315 | 50,35.84 316 | 10,18.83 317 | 20,10.08 318 | 30,22.71 319 | 40,26.29 320 | 50,29.64 321 | 10,25.72 322 | 20,13.61 323 | 30,33.18 324 | 40,33.39 325 | 50,37.29 326 | 10,16.81 327 | 20,23.41 328 | 30,23.46 329 | 40,34.31 330 | 50,43.47 331 | 10,10.28 332 | 20,20.32 333 | 30,22.98 334 | 40,34.78 335 | 50,34.75 336 | 10,17.38 337 | 20,16.84 338 | 30,27.91 339 | 40,29.32 340 | 50,36.58 341 | 10,12.35 342 | 20,25.81 343 | 30,22.79 344 | 40,28.24 345 | 50,31.52 346 | 10,18.03 347 | 20,12.55 348 | 30,18.16 349 | 40,28.44 350 | 50,32.55 351 | 10,10.27 352 | 20,20.78 353 | 30,22.81 354 | 40,32.64 355 | 50,30.21 356 | 10,20.37 357 | 20,21.41 358 | 30,19.16 359 | 40,21.83 360 | 50,36.12 361 | 10,21.42 362 | 20,16.28 363 | 30,28.19 364 | 40,27.73 365 | 50,33.33 366 | 10,21.53 367 | 20,24.48 368 | 30,28.21 369 | 40,35.43 370 | 50,34.80 371 | 10,26.89 372 | 20,20.11 373 | 30,27.01 374 | 40,28.83 375 | 50,30.15 376 | 10,14.90 377 | 20,17.34 378 | 30,29.03 379 | 40,30.09 380 | 50,41.26 381 | 10,14.23 382 | 20,20.17 383 | 30,27.35 384 | 40,24.21 385 | 50,25.74 386 | 10,23.08 387 | 20,13.31 388 | 30,30.77 389 | 40,36.52 390 | 50,38.26 391 | 10,15.66 392 | 20,21.70 393 | 30,15.13 394 | 40,37.97 395 | 50,30.13 396 | 10,14.85 397 | 20,19.17 398 | 30,30.22 399 | 40,31.50 400 | 50,32.20 401 | 10,16.54 402 | 20,16.03 403 | 30,25.67 404 | 40,31.65 405 | 50,26.43 406 | 10,9.75 407 | 20,19.10 408 | 30,32.51 409 | 40,28.51 410 | 50,32.62 411 | 10,14.64 412 | 20,18.05 413 | 30,18.31 414 | 40,31.17 415 | 50,41.50 416 | 10,22.82 417 | 20,25.76 418 | 30,35.16 419 | 40,33.44 420 | 50,36.65 421 | 10,11.05 422 | 20,19.71 423 | 30,14.27 424 | 40,34.11 425 | 50,38.20 426 | 10,18.10 427 | 20,21.80 428 | 30,28.28 429 | 40,34.97 430 | 50,36.24 431 | 10,15.32 432 | 20,23.06 433 | 30,27.65 434 | 40,33.32 435 | 50,42.25 436 | 10,21.16 437 | 20,25.56 438 | 30,30.47 439 | 40,34.34 440 | 50,44.66 441 | 10,13.40 442 | 20,15.19 443 | 30,25.34 444 | 40,41.86 445 | 50,32.46 446 | 10,19.57 447 | 20,15.92 448 | 30,26.20 449 | 40,27.70 450 | 50,38.89 451 | 10,13.22 452 | 20,26.76 453 | 30,24.59 454 | 40,30.85 455 | 50,35.20 456 | 10,12.23 457 | 20,20.56 458 | 30,24.73 459 | 40,39.86 460 | 50,33.60 461 | 10,7.08 462 | 20,16.36 463 | 30,29.65 464 | 40,31.37 465 | 50,32.40 466 | 10,15.32 467 | 20,20.10 468 | 30,28.50 469 | 40,24.39 470 | 50,26.86 471 | 10,8.96 472 | 20,19.93 473 | 30,28.13 474 | 40,29.01 475 | 50,36.56 476 | 10,11.12 477 | 20,10.80 478 | 30,30.45 479 | 40,22.08 480 | 50,34.36 481 | 10,18.49 482 | 20,9.70 483 | 30,36.77 484 | 40,24.12 485 | 50,37.62 486 | 10,18.53 487 | 20,19.76 488 | 30,23.09 489 | 40,38.86 490 | 50,28.41 491 | 10,15.54 492 | 20,27.79 493 | 30,17.89 494 | 40,25.83 495 | 50,36.29 496 | 10,10.25 497 | 20,22.74 498 | 30,26.71 499 | 40,35.93 500 | 50,33.16 501 | -------------------------------------------------------------------------------- /ResultsFromPaper/FashionMNIST_vae_k10_p10laplace_none_accuracy.csv: -------------------------------------------------------------------------------- 1 | 10,50.07 2 | 20,61.84 3 | 30,68.65 4 | 40,68.01 5 | 50,68.67 6 | 10,56.82 7 | 20,56.96 8 | 30,66.04 9 | 40,74.77 10 | 50,68.06 11 | 10,61.59 12 | 20,52.98 13 | 30,64.15 14 | 40,67.04 15 | 50,67.63 16 | 10,55.28 17 | 20,57.34 18 | 30,62.01 19 | 40,68.00 20 | 50,69.62 21 | 10,59.45 22 | 20,58.87 23 | 30,67.88 24 | 40,69.70 25 | 50,68.43 26 | 10,54.43 27 | 20,58.30 28 | 30,67.10 29 | 40,69.52 30 | 50,68.97 31 | 10,55.02 32 | 20,58.65 33 | 30,65.72 34 | 40,65.87 35 | 50,68.42 36 | 10,47.06 37 | 20,63.96 38 | 30,59.17 39 | 40,68.18 40 | 50,68.55 41 | 10,53.84 42 | 20,49.97 43 | 30,65.51 44 | 40,65.81 45 | 50,68.13 46 | 10,52.14 47 | 20,55.45 48 | 30,68.02 49 | 40,63.13 50 | 50,68.00 51 | -------------------------------------------------------------------------------- /ResultsFromPaper/FashionMNIST_vae_k10_poisson_accuracy50trials.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-08_14:36 2 | 50,71.66 3 | 20,63.66 4 | 50,70.39 5 | 40,76.74 6 | 30,71.87 7 | 10,58.27 8 | 30,70.03 9 | 40,70.66 10 | 10,62.37 11 | 20,62.91 12 | 20,62.92 13 | 30,68.27 14 | 10,66.94 15 | 50,69.07 16 | 10,60.00 17 | 40,70.96 18 | 20,65.48 19 | 10,66.41 20 | 40,73.71 21 | 30,66.57 22 | 50,72.17 23 | 20,66.89 24 | 40,71.39 25 | 10,64.21 26 | 30,71.69 27 | 40,75.50 28 | 20,66.83 29 | 50,72.60 30 | 30,71.41 31 | 30,71.00 32 | 20,63.88 33 | 50,72.69 34 | 50,74.39 35 | 20,66.45 36 | 40,71.62 37 | 10,60.85 38 | 40,70.50 39 | 10,56.46 40 | 50,70.82 41 | 50,72.98 42 | 20,58.94 43 | 30,60.41 44 | 30,68.04 45 | 40,71.13 46 | 10,59.80 47 | 10,61.11 48 | 20,65.96 49 | 40,65.78 50 | 50,71.36 51 | 30,70.86 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/FashionMNIST_vae_k10_sparselabelpropagation_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-02_14:29 2 | 30,12.23 3 | 40,20.73 4 | 40,14.27 5 | 10,6.52 6 | 30,17.15 7 | 50,12.03 8 | 50,16.85 9 | 20,13.33 10 | 10,21.31 11 | 20,16.32 12 | 10,17.93 13 | 20,17.05 14 | 40,17.59 15 | 40,13.22 16 | 30,16.24 17 | 50,19.11 18 | 50,14.82 19 | 30,13.67 20 | 10,12.28 21 | 20,15.76 22 | 10,13.74 23 | 20,16.76 24 | 40,7.92 25 | 30,14.72 26 | 40,11.12 27 | 50,15.88 28 | 10,10.91 29 | 30,14.92 30 | 20,14.67 31 | 50,17.50 32 | 10,14.83 33 | 30,15.07 34 | 40,14.05 35 | 20,15.14 36 | 50,20.66 37 | 40,14.23 38 | 10,14.24 39 | 30,4.61 40 | 20,16.90 41 | 50,13.38 42 | 20,18.66 43 | 10,15.62 44 | 30,14.48 45 | 10,13.59 46 | 40,11.29 47 | 50,14.27 48 | 20,20.88 49 | 30,13.85 50 | 40,13.89 51 | 50,16.22 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/FashionMNIST_vae_k10_wnll_accuracy.csv: -------------------------------------------------------------------------------- 1 | 10,40.74 2 | 20,62.58 3 | 30,64.21 4 | 40,67.46 5 | 50,68.23 6 | 10,43.98 7 | 20,58.54 8 | 30,65.84 9 | 40,73.77 10 | 50,67.17 11 | 10,59.35 12 | 20,55.14 13 | 30,66.25 14 | 40,68.47 15 | 50,70.44 16 | 10,47.08 17 | 20,57.78 18 | 30,56.73 19 | 40,69.33 20 | 50,67.10 21 | 10,50.92 22 | 20,66.38 23 | 30,69.82 24 | 40,66.44 25 | 50,73.16 26 | 10,49.67 27 | 20,59.68 28 | 30,66.32 29 | 40,64.90 30 | 50,72.87 31 | 10,46.00 32 | 20,54.88 33 | 30,64.95 34 | 40,64.70 35 | 50,70.64 36 | 10,46.54 37 | 20,64.98 38 | 30,59.14 39 | 40,69.01 40 | 50,68.32 41 | 10,38.37 42 | 20,51.19 43 | 30,63.68 44 | 40,67.66 45 | 50,69.56 46 | 10,49.27 47 | 20,56.15 48 | 30,63.32 49 | 40,63.31 50 | 50,70.02 51 | 10,43.42 52 | 20,67.67 53 | 30,65.35 54 | 40,70.06 55 | 50,73.24 56 | 10,37.02 57 | 20,59.11 58 | 30,64.21 59 | 40,60.93 60 | 50,71.31 61 | 10,42.72 62 | 20,62.50 63 | 30,65.04 64 | 40,63.42 65 | 50,68.59 66 | 10,44.53 67 | 20,64.17 68 | 30,62.25 69 | 40,67.22 70 | 50,65.78 71 | 10,32.03 72 | 20,58.70 73 | 30,65.62 74 | 40,70.23 75 | 50,69.60 76 | 10,40.06 77 | 20,64.20 78 | 30,68.05 79 | 40,68.32 80 | 50,73.75 81 | 10,39.56 82 | 20,50.86 83 | 30,63.40 84 | 40,70.45 85 | 50,71.76 86 | 10,30.58 87 | 20,59.61 88 | 30,65.16 89 | 40,67.06 90 | 50,72.05 91 | 10,49.05 92 | 20,59.87 93 | 30,68.48 94 | 40,64.19 95 | 50,67.05 96 | 10,39.39 97 | 20,68.19 98 | 30,66.17 99 | 40,68.27 100 | 50,70.77 101 | 10,61.17 102 | 20,58.57 103 | 30,59.07 104 | 40,65.97 105 | 50,63.88 106 | 10,48.36 107 | 20,51.79 108 | 30,66.89 109 | 40,69.19 110 | 50,72.45 111 | 10,37.38 112 | 20,52.44 113 | 30,61.11 114 | 40,67.04 115 | 50,67.73 116 | 10,46.51 117 | 20,53.29 118 | 30,63.85 119 | 40,66.36 120 | 50,70.10 121 | 10,50.66 122 | 20,57.49 123 | 30,61.15 124 | 40,64.93 125 | 50,68.22 126 | 10,34.69 127 | 20,59.84 128 | 30,68.00 129 | 40,67.71 130 | 50,61.00 131 | 10,31.88 132 | 20,55.39 133 | 30,62.05 134 | 40,64.67 135 | 50,68.94 136 | 10,46.78 137 | 20,64.96 138 | 30,61.41 139 | 40,68.85 140 | 50,70.13 141 | 10,50.93 142 | 20,58.66 143 | 30,63.38 144 | 40,66.27 145 | 50,70.71 146 | 10,34.06 147 | 20,63.80 148 | 30,67.98 149 | 40,69.51 150 | 50,68.07 151 | 10,45.41 152 | 20,60.41 153 | 30,68.73 154 | 40,69.12 155 | 50,71.43 156 | 10,43.80 157 | 20,65.69 158 | 30,62.68 159 | 40,69.08 160 | 50,69.30 161 | 10,40.99 162 | 20,62.14 163 | 30,64.29 164 | 40,68.04 165 | 50,71.39 166 | 10,42.05 167 | 20,60.67 168 | 30,55.43 169 | 40,70.04 170 | 50,73.66 171 | 10,43.61 172 | 20,66.58 173 | 30,65.97 174 | 40,65.30 175 | 50,64.75 176 | 10,44.78 177 | 20,60.15 178 | 30,69.66 179 | 40,64.37 180 | 50,71.57 181 | 10,28.27 182 | 20,62.69 183 | 30,57.95 184 | 40,67.13 185 | 50,68.40 186 | 10,42.16 187 | 20,55.94 188 | 30,68.57 189 | 40,69.58 190 | 50,69.26 191 | 10,51.59 192 | 20,55.49 193 | 30,70.21 194 | 40,70.58 195 | 50,74.92 196 | 10,41.82 197 | 20,61.99 198 | 30,63.32 199 | 40,65.45 200 | 50,68.49 201 | 10,51.58 202 | 20,61.54 203 | 30,64.99 204 | 40,73.48 205 | 50,66.51 206 | 10,50.51 207 | 20,50.19 208 | 30,61.29 209 | 40,69.00 210 | 50,67.62 211 | 10,54.41 212 | 20,51.73 213 | 30,63.87 214 | 40,65.21 215 | 50,67.43 216 | 10,50.91 217 | 20,61.12 218 | 30,67.26 219 | 40,68.54 220 | 50,71.17 221 | 10,51.71 222 | 20,58.34 223 | 30,60.17 224 | 40,61.99 225 | 50,68.84 226 | 10,46.48 227 | 20,55.57 228 | 30,63.13 229 | 40,66.30 230 | 50,73.41 231 | 10,58.04 232 | 20,57.41 233 | 30,66.98 234 | 40,64.21 235 | 50,72.94 236 | 10,57.79 237 | 20,53.94 238 | 30,60.52 239 | 40,69.20 240 | 50,74.26 241 | 10,52.25 242 | 20,59.78 243 | 30,57.84 244 | 40,70.25 245 | 50,66.54 246 | 10,31.35 247 | 20,59.29 248 | 30,69.15 249 | 40,69.74 250 | 50,70.15 251 | 10,45.03 252 | 20,54.58 253 | 30,66.18 254 | 40,68.44 255 | 50,67.33 256 | 10,40.60 257 | 20,63.03 258 | 30,64.48 259 | 40,64.02 260 | 50,68.37 261 | 10,39.40 262 | 20,57.54 263 | 30,63.62 264 | 40,63.43 265 | 50,73.62 266 | 10,46.30 267 | 20,64.25 268 | 30,68.10 269 | 40,62.30 270 | 50,75.40 271 | 10,43.09 272 | 20,60.81 273 | 30,60.62 274 | 40,70.65 275 | 50,67.13 276 | 10,55.20 277 | 20,58.19 278 | 30,70.22 279 | 40,64.35 280 | 50,69.68 281 | 10,41.23 282 | 20,58.43 283 | 30,67.55 284 | 40,68.03 285 | 50,70.15 286 | 10,38.71 287 | 20,63.80 288 | 30,66.78 289 | 40,68.34 290 | 50,66.42 291 | 10,54.04 292 | 20,56.85 293 | 30,67.14 294 | 40,70.02 295 | 50,67.61 296 | 10,45.64 297 | 20,45.64 298 | 30,65.56 299 | 40,68.93 300 | 50,73.02 301 | 10,44.37 302 | 20,49.41 303 | 30,62.31 304 | 40,58.91 305 | 50,71.83 306 | 10,45.48 307 | 20,60.10 308 | 30,62.44 309 | 40,66.07 310 | 50,69.31 311 | 10,44.95 312 | 20,62.01 313 | 30,66.07 314 | 40,65.35 315 | 50,66.60 316 | 10,51.27 317 | 20,53.17 318 | 30,71.33 319 | 40,70.22 320 | 50,65.40 321 | 10,49.57 322 | 20,54.25 323 | 30,68.96 324 | 40,66.05 325 | 50,71.26 326 | 10,50.81 327 | 20,56.00 328 | 30,65.97 329 | 40,72.27 330 | 50,67.64 331 | 10,40.64 332 | 20,59.76 333 | 30,66.28 334 | 40,70.70 335 | 50,71.23 336 | 10,53.28 337 | 20,64.97 338 | 30,65.94 339 | 40,70.44 340 | 50,73.13 341 | 10,42.19 342 | 20,54.35 343 | 30,60.17 344 | 40,69.39 345 | 50,69.76 346 | 10,44.59 347 | 20,62.59 348 | 30,67.96 349 | 40,65.78 350 | 50,72.40 351 | 10,43.05 352 | 20,61.75 353 | 30,61.82 354 | 40,71.23 355 | 50,72.43 356 | 10,43.65 357 | 20,63.20 358 | 30,69.62 359 | 40,68.37 360 | 50,67.04 361 | 10,34.63 362 | 20,44.78 363 | 30,66.10 364 | 40,69.37 365 | 50,70.81 366 | 10,41.87 367 | 20,60.14 368 | 30,61.79 369 | 40,65.99 370 | 50,72.03 371 | 10,49.83 372 | 20,59.32 373 | 30,62.27 374 | 40,71.34 375 | 50,68.68 376 | 10,27.91 377 | 20,68.56 378 | 30,61.71 379 | 40,65.14 380 | 50,71.45 381 | 10,49.23 382 | 20,60.73 383 | 30,64.26 384 | 40,66.06 385 | 50,69.31 386 | 10,47.86 387 | 20,55.80 388 | 30,68.01 389 | 40,63.70 390 | 50,72.96 391 | 10,40.61 392 | 20,55.85 393 | 30,63.04 394 | 40,70.59 395 | 50,67.16 396 | 10,38.39 397 | 20,58.38 398 | 30,72.46 399 | 40,66.12 400 | 50,73.92 401 | 10,43.95 402 | 20,65.20 403 | 30,62.66 404 | 40,68.00 405 | 50,73.52 406 | 10,42.87 407 | 20,59.93 408 | 30,62.92 409 | 40,71.18 410 | 50,67.83 411 | 10,44.82 412 | 20,56.44 413 | 30,67.72 414 | 40,66.14 415 | 50,75.41 416 | 10,52.50 417 | 20,58.43 418 | 30,63.31 419 | 40,62.51 420 | 50,69.35 421 | 10,47.82 422 | 20,64.16 423 | 30,56.03 424 | 40,63.46 425 | 50,69.70 426 | 10,28.57 427 | 20,53.28 428 | 30,72.59 429 | 40,66.53 430 | 50,69.64 431 | 10,34.86 432 | 20,63.95 433 | 30,70.88 434 | 40,72.44 435 | 50,70.74 436 | 10,44.18 437 | 20,60.00 438 | 30,61.63 439 | 40,65.10 440 | 50,73.43 441 | 10,52.78 442 | 20,55.59 443 | 30,65.93 444 | 40,69.91 445 | 50,67.75 446 | 10,44.32 447 | 20,60.99 448 | 30,64.42 449 | 40,67.93 450 | 50,70.11 451 | 10,42.09 452 | 20,55.54 453 | 30,64.69 454 | 40,63.18 455 | 50,72.73 456 | 10,44.43 457 | 20,53.27 458 | 30,67.33 459 | 40,67.86 460 | 50,75.68 461 | 10,39.58 462 | 20,63.26 463 | 30,63.50 464 | 40,72.57 465 | 50,67.60 466 | 10,50.79 467 | 20,62.11 468 | 30,63.12 469 | 40,69.57 470 | 50,64.85 471 | 10,53.34 472 | 20,59.64 473 | 30,64.29 474 | 40,70.26 475 | 50,70.41 476 | 10,48.75 477 | 20,63.64 478 | 30,68.60 479 | 40,50.43 480 | 50,71.14 481 | 10,29.42 482 | 20,61.73 483 | 30,66.35 484 | 40,70.41 485 | 50,72.50 486 | 10,56.08 487 | 20,58.67 488 | 30,62.83 489 | 40,66.55 490 | 50,69.96 491 | 10,37.49 492 | 20,67.63 493 | 30,61.37 494 | 40,69.24 495 | 50,74.72 496 | 10,49.37 497 | 20,59.30 498 | 30,62.05 499 | 40,71.29 500 | 50,68.13 501 | -------------------------------------------------------------------------------- /ResultsFromPaper/FashionMNIST_vae_k12_poisson_accuracy50trials.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-05_06:45 2 | 10,62.43 3 | 40,70.78 4 | 50,71.72 5 | 40,77.15 6 | 50,70.50 7 | 20,63.81 8 | 10,58.16 9 | 20,62.87 10 | 30,70.17 11 | 10,67.65 12 | 30,71.98 13 | 20,62.79 14 | 50,69.37 15 | 30,66.86 16 | 50,72.39 17 | 30,68.46 18 | 40,71.23 19 | 10,59.98 20 | 40,71.45 21 | 40,73.95 22 | 10,66.45 23 | 20,65.27 24 | 50,72.99 25 | 20,66.88 26 | 10,64.02 27 | 20,66.90 28 | 30,71.60 29 | 40,75.69 30 | 40,71.64 31 | 30,70.91 32 | 30,71.56 33 | 20,63.75 34 | 50,72.97 35 | 50,74.47 36 | 40,70.33 37 | 10,60.78 38 | 20,66.34 39 | 10,56.21 40 | 50,70.94 41 | 30,68.10 42 | 50,73.09 43 | 10,61.56 44 | 30,60.95 45 | 20,58.99 46 | 50,71.45 47 | 40,71.35 48 | 40,65.85 49 | 10,60.11 50 | 20,66.47 51 | 30,71.02 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/FashionMNIST_vae_k14_poisson_accuracy50trials.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-05_06:50 2 | 30,72.01 3 | 30,70.10 4 | 10,58.68 5 | 50,71.71 6 | 40,70.93 7 | 10,61.89 8 | 40,77.24 9 | 50,70.71 10 | 10,67.62 11 | 20,63.66 12 | 20,62.67 13 | 20,62.77 14 | 40,74.06 15 | 50,69.46 16 | 30,68.67 17 | 10,59.88 18 | 40,71.33 19 | 50,72.34 20 | 10,66.12 21 | 30,71.35 22 | 30,66.84 23 | 10,64.05 24 | 20,65.42 25 | 20,66.93 26 | 50,74.37 27 | 20,66.37 28 | 50,73.42 29 | 40,71.57 30 | 20,63.49 31 | 40,75.74 32 | 30,70.74 33 | 40,70.47 34 | 50,73.14 35 | 10,60.30 36 | 40,71.51 37 | 20,66.47 38 | 30,71.65 39 | 30,68.40 40 | 10,55.98 41 | 40,71.35 42 | 50,73.16 43 | 30,61.34 44 | 20,59.08 45 | 10,61.83 46 | 50,70.98 47 | 50,71.38 48 | 10,60.19 49 | 20,66.56 50 | 40,65.84 51 | 30,71.03 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/FashionMNIST_vae_k16_poisson_accuracy50trials.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-08_13:39 2 | 10,58.93 3 | 30,72.00 4 | 40,71.03 5 | 20,62.68 6 | 10,67.49 7 | 50,71.71 8 | 10,61.69 9 | 20,62.50 10 | 50,70.77 11 | 40,77.04 12 | 20,63.70 13 | 50,69.57 14 | 30,70.03 15 | 30,68.68 16 | 30,66.95 17 | 40,71.41 18 | 10,65.56 19 | 20,65.80 20 | 50,73.58 21 | 50,72.40 22 | 40,74.00 23 | 30,71.18 24 | 10,59.85 25 | 30,70.70 26 | 10,64.01 27 | 20,66.86 28 | 40,75.62 29 | 20,66.07 30 | 40,71.58 31 | 40,71.40 32 | 50,74.41 33 | 20,66.66 34 | 50,73.07 35 | 10,60.07 36 | 20,63.52 37 | 50,73.17 38 | 40,70.43 39 | 30,71.74 40 | 30,62.02 41 | 20,58.91 42 | 10,56.08 43 | 30,68.65 44 | 50,70.97 45 | 10,59.79 46 | 50,71.36 47 | 10,61.82 48 | 40,71.40 49 | 40,65.82 50 | 20,66.59 51 | 30,70.96 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/FashionMNIST_vae_k18_poisson_accuracy50trials.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-05_06:54 2 | 10,61.51 3 | 20,63.73 4 | 10,58.73 5 | 40,71.06 6 | 30,72.03 7 | 50,71.73 8 | 20,62.61 9 | 10,67.73 10 | 20,62.40 11 | 50,70.93 12 | 40,77.09 13 | 50,69.61 14 | 30,70.11 15 | 30,68.78 16 | 40,74.12 17 | 40,71.42 18 | 10,65.46 19 | 30,66.83 20 | 50,72.40 21 | 50,73.55 22 | 20,65.93 23 | 10,59.83 24 | 30,71.15 25 | 20,66.89 26 | 20,65.89 27 | 50,74.42 28 | 40,71.49 29 | 10,60.03 30 | 40,75.60 31 | 10,64.02 32 | 30,70.68 33 | 30,61.88 34 | 10,56.09 35 | 40,71.29 36 | 20,63.35 37 | 50,73.10 38 | 20,66.74 39 | 30,71.72 40 | 40,70.41 41 | 20,58.89 42 | 30,68.76 43 | 50,73.11 44 | 40,71.28 45 | 10,61.89 46 | 50,70.92 47 | 50,71.42 48 | 10,59.74 49 | 40,65.81 50 | 20,66.52 51 | 30,70.98 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/FashionMNIST_vae_k20_poisson_accuracy50trials.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-05_06:58 2 | 10,58.54 3 | 10,61.63 4 | 30,72.08 5 | 50,71.79 6 | 40,71.01 7 | 10,67.62 8 | 20,62.59 9 | 20,62.42 10 | 40,77.12 11 | 50,70.88 12 | 20,63.72 13 | 10,65.33 14 | 30,70.16 15 | 30,68.81 16 | 50,69.54 17 | 30,66.94 18 | 20,66.74 19 | 40,71.43 20 | 20,66.23 21 | 40,74.29 22 | 50,72.36 23 | 10,63.99 24 | 10,59.78 25 | 40,75.59 26 | 30,71.18 27 | 50,73.51 28 | 20,65.95 29 | 40,71.44 30 | 40,71.17 31 | 30,70.59 32 | 20,63.25 33 | 10,59.71 34 | 50,74.40 35 | 30,62.06 36 | 40,70.36 37 | 20,66.75 38 | 50,73.11 39 | 10,61.87 40 | 30,71.68 41 | 50,73.10 42 | 10,55.89 43 | 40,71.30 44 | 20,58.99 45 | 30,68.79 46 | 50,70.89 47 | 50,71.43 48 | 20,66.53 49 | 30,70.97 50 | 40,65.84 51 | 10,59.49 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/FashionMNIST_vae_k6_poisson_accuracy50trials.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-05_06:34 2 | 10,57.86 3 | 50,71.01 4 | 30,71.66 5 | 50,69.62 6 | 40,75.75 7 | 10,63.41 8 | 40,70.53 9 | 20,62.76 10 | 10,65.14 11 | 30,70.29 12 | 20,62.75 13 | 20,62.17 14 | 50,68.14 15 | 10,66.46 16 | 40,73.12 17 | 40,69.79 18 | 30,67.67 19 | 30,65.91 20 | 50,72.17 21 | 10,59.95 22 | 30,71.38 23 | 20,62.79 24 | 40,69.70 25 | 20,66.30 26 | 10,64.23 27 | 10,60.91 28 | 50,72.03 29 | 20,66.99 30 | 30,70.83 31 | 30,70.45 32 | 50,74.06 33 | 40,74.47 34 | 20,63.42 35 | 10,55.73 36 | 40,71.48 37 | 40,70.45 38 | 50,72.20 39 | 50,72.68 40 | 20,65.67 41 | 30,59.26 42 | 10,59.96 43 | 20,58.42 44 | 30,68.03 45 | 40,70.44 46 | 10,60.06 47 | 50,70.44 48 | 30,70.45 49 | 20,63.39 50 | 40,65.50 51 | 50,71.05 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/FashionMNIST_vae_k8_poisson_accuracy50trials.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-05_06:41 2 | 10,62.08 3 | 20,63.65 4 | 50,71.70 5 | 10,59.53 6 | 40,76.50 7 | 50,70.28 8 | 40,70.79 9 | 30,71.62 10 | 10,67.11 11 | 30,70.13 12 | 20,63.16 13 | 50,69.49 14 | 40,73.86 15 | 20,62.94 16 | 30,68.56 17 | 10,59.79 18 | 40,71.05 19 | 10,66.20 20 | 30,66.46 21 | 10,64.47 22 | 30,71.64 23 | 50,72.45 24 | 50,72.79 25 | 20,66.89 26 | 40,75.57 27 | 20,65.01 28 | 50,74.35 29 | 30,70.83 30 | 20,63.56 31 | 10,60.72 32 | 40,70.95 33 | 50,72.74 34 | 20,66.56 35 | 40,70.52 36 | 30,71.30 37 | 20,58.91 38 | 40,71.10 39 | 30,60.15 40 | 50,73.04 41 | 20,66.63 42 | 40,71.45 43 | 50,70.77 44 | 20,65.59 45 | 30,68.31 46 | 10,56.40 47 | 40,65.81 48 | 10,59.63 49 | 30,70.97 50 | 50,71.30 51 | 10,60.79 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/MNISTModRate_vae_k10_p10laplace_none_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-03_14:13 2 | 1600,96.22 3 | 1600,96.42 4 | 800,96.29 5 | 800,96.10 6 | 400,95.45 7 | 400,95.48 8 | 200,94.46 9 | 200,94.78 10 | 800,96.09 11 | 1600,96.04 12 | 100,93.36 13 | 400,95.24 14 | 100,94.09 15 | 200,95.51 16 | 1600,96.29 17 | 100,94.49 18 | 800,96.09 19 | 400,95.46 20 | 200,95.51 21 | 100,94.54 22 | 1600,96.29 23 | 800,95.98 24 | 400,95.58 25 | 200,95.34 26 | 1600,96.36 27 | 400,95.60 28 | 800,96.12 29 | 200,94.29 30 | 100,93.50 31 | 100,94.10 32 | 400,95.59 33 | 1600,95.93 34 | 200,95.34 35 | 800,96.03 36 | 100,94.21 37 | 1600,96.17 38 | 800,95.93 39 | 400,95.55 40 | 200,95.09 41 | 1600,96.07 42 | 800,95.86 43 | 400,95.45 44 | 200,95.20 45 | 100,94.32 46 | 1600,96.17 47 | 100,94.82 48 | 400,95.40 49 | 800,95.61 50 | 200,95.22 51 | 100,92.07 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/MNISTModRate_vae_k10_sparselabelpropagation_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-03_12:03 2 | 400,23.42 3 | 100,19.04 4 | 100,19.52 5 | 200,18.87 6 | 1600,58.90 7 | 800,30.17 8 | 200,21.24 9 | 1600,53.21 10 | 400,24.50 11 | 800,20.64 12 | 100,23.21 13 | 400,22.92 14 | 200,15.07 15 | 200,21.29 16 | 1600,49.49 17 | 800,21.50 18 | 1600,49.77 19 | 100,11.39 20 | 400,21.43 21 | 800,38.33 22 | 100,21.54 23 | 200,22.90 24 | 800,32.20 25 | 400,22.02 26 | 100,18.93 27 | 200,19.91 28 | 800,36.24 29 | 1600,58.60 30 | 400,23.74 31 | 1600,49.48 32 | 200,18.02 33 | 400,21.42 34 | 100,20.23 35 | 800,37.36 36 | 100,18.61 37 | 1600,50.04 38 | 1600,58.93 39 | 800,29.52 40 | 400,21.83 41 | 200,24.57 42 | 200,21.15 43 | 400,22.82 44 | 100,19.26 45 | 800,29.67 46 | 800,31.63 47 | 200,23.24 48 | 1600,58.92 49 | 100,18.10 50 | 400,22.47 51 | 1600,58.59 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/MNISTMult1_vae_k10_modularitymbo_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2021-03-22_23:02 2 | 10,88.11 3 | 10,94.81 4 | 10,93.69 5 | 10,91.82 6 | 10,90.72 7 | 10,95.42 8 | 10,96.05 9 | 10,95.94 10 | 10,96.03 11 | 10,95.41 12 | 10,92.53 13 | 10,95.71 14 | 10,92.24 15 | 10,95.34 16 | 10,95.40 17 | 10,92.21 18 | 10,92.55 19 | 10,96.39 20 | 10,96.93 21 | 10,97.12 22 | 10,95.93 23 | 10,95.38 24 | 10,94.25 25 | 10,95.99 26 | 10,95.96 27 | 10,81.96 28 | 10,91.83 29 | 10,92.14 30 | 10,86.71 31 | 10,75.33 32 | 10,95.33 33 | 10,91.92 34 | 10,92.34 35 | 10,91.79 36 | 10,96.11 37 | 10,96.01 38 | 10,96.40 39 | 10,96.96 40 | 10,96.91 41 | 10,91.42 42 | 10,96.90 43 | 10,83.11 44 | 10,90.66 45 | 10,95.98 46 | 10,91.92 47 | 10,93.11 48 | 10,96.05 49 | 10,96.39 50 | 10,92.62 51 | 10,93.20 52 | 10,96.04 53 | 10,89.77 54 | 10,78.32 55 | 10,95.35 56 | 10,96.08 57 | 10,96.91 58 | 10,93.60 59 | 10,96.92 60 | 10,96.11 61 | 10,73.83 62 | 10,95.39 63 | 10,87.25 64 | 10,90.69 65 | 10,90.15 66 | 10,95.39 67 | 10,95.95 68 | 10,87.02 69 | 10,96.58 70 | 10,95.43 71 | 10,94.84 72 | 10,95.99 73 | 10,95.38 74 | 10,95.97 75 | 10,86.33 76 | 10,92.39 77 | 10,96.92 78 | 10,92.39 79 | 10,83.20 80 | 10,95.97 81 | 10,96.00 82 | 10,96.08 83 | 10,96.01 84 | 10,91.18 85 | 10,95.93 86 | 10,92.48 87 | 10,89.94 88 | 10,95.96 89 | 10,95.43 90 | 10,83.33 91 | 10,92.08 92 | 10,96.00 93 | 10,95.49 94 | 10,92.71 95 | 10,93.43 96 | 10,96.26 97 | 10,92.21 98 | 10,88.04 99 | 10,96.91 100 | 10,95.72 101 | 10,95.97 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/MNISTMult1_vae_k10_poisson_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-08_22:28 2 | 10,91.82 3 | 10,87.02 4 | 10,84.63 5 | 10,92.56 6 | 10,92.90 7 | 10,88.17 8 | 10,94.22 9 | 10,93.69 10 | 10,92.34 11 | 10,89.73 12 | 10,92.85 13 | 10,92.39 14 | 10,93.71 15 | 10,93.21 16 | 10,93.36 17 | 10,94.05 18 | 10,95.44 19 | 10,80.50 20 | 10,91.67 21 | 10,91.10 22 | 10,93.56 23 | 10,90.38 24 | 10,94.87 25 | 10,92.44 26 | 10,93.65 27 | 10,93.51 28 | 10,89.36 29 | 10,90.05 30 | 10,90.11 31 | 10,92.94 32 | 10,93.43 33 | 10,81.09 34 | 10,92.42 35 | 10,86.88 36 | 10,91.14 37 | 10,91.79 38 | 10,89.08 39 | 10,95.47 40 | 10,94.25 41 | 10,92.58 42 | 10,91.55 43 | 10,78.43 44 | 10,88.73 45 | 10,88.62 46 | 10,94.93 47 | 10,92.38 48 | 10,94.34 49 | 10,91.12 50 | 10,89.16 51 | 10,90.10 52 | 10,79.97 53 | 10,94.44 54 | 10,87.83 55 | 10,84.17 56 | 10,89.04 57 | 10,94.07 58 | 10,91.67 59 | 10,92.72 60 | 10,75.96 61 | 10,91.56 62 | 10,90.71 63 | 10,94.81 64 | 10,83.63 65 | 10,93.31 66 | 10,88.41 67 | 10,91.17 68 | 10,93.49 69 | 10,92.90 70 | 10,92.48 71 | 10,94.60 72 | 10,94.29 73 | 10,90.15 74 | 10,89.82 75 | 10,91.16 76 | 10,92.23 77 | 10,89.09 78 | 10,89.74 79 | 10,95.10 80 | 10,95.42 81 | 10,92.60 82 | 10,92.04 83 | 10,92.74 84 | 10,90.13 85 | 10,80.04 86 | 10,87.67 87 | 10,92.88 88 | 10,92.98 89 | 10,93.77 90 | 10,79.13 91 | 10,84.35 92 | 10,93.30 93 | 10,91.64 94 | 10,93.02 95 | 10,94.23 96 | 10,93.48 97 | 10,94.41 98 | 10,94.10 99 | 10,94.65 100 | 10,92.09 101 | 10,89.62 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/MNISTMult2_vae_k10_modularitymbo_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2021-03-22_23:02 2 | 15,92.13 3 | 15,96.35 4 | 15,96.32 5 | 15,96.39 6 | 15,95.95 7 | 15,92.26 8 | 15,92.12 9 | 15,92.08 10 | 15,92.49 11 | 15,88.60 12 | 15,96.93 13 | 15,92.17 14 | 15,96.15 15 | 15,92.46 16 | 15,91.92 17 | 15,95.53 18 | 15,96.39 19 | 15,95.94 20 | 15,96.23 21 | 15,96.95 22 | 15,95.51 23 | 15,94.46 24 | 15,96.05 25 | 15,95.93 26 | 15,93.10 27 | 15,96.01 28 | 15,96.10 29 | 15,95.93 30 | 15,89.86 31 | 15,95.42 32 | 15,96.37 33 | 15,83.44 34 | 15,95.97 35 | 15,92.42 36 | 15,95.97 37 | 15,96.29 38 | 15,92.28 39 | 15,92.58 40 | 15,92.46 41 | 15,95.94 42 | 15,95.02 43 | 15,95.37 44 | 15,92.83 45 | 15,95.53 46 | 15,94.83 47 | 15,96.90 48 | 15,96.91 49 | 15,95.33 50 | 15,92.89 51 | 15,68.73 52 | 15,95.90 53 | 15,96.36 54 | 15,95.44 55 | 15,96.94 56 | 15,82.71 57 | 15,90.60 58 | 15,95.51 59 | 15,96.93 60 | 15,92.96 61 | 15,95.40 62 | 15,96.00 63 | 15,95.94 64 | 15,95.98 65 | 15,92.06 66 | 15,95.50 67 | 15,95.43 68 | 15,95.94 69 | 15,96.91 70 | 15,95.30 71 | 15,84.54 72 | 15,79.50 73 | 15,92.25 74 | 15,91.20 75 | 15,95.96 76 | 15,95.98 77 | 15,95.97 78 | 15,96.91 79 | 15,95.95 80 | 15,96.08 81 | 15,95.93 82 | 15,95.99 83 | 15,95.55 84 | 15,96.03 85 | 15,80.18 86 | 15,96.36 87 | 15,95.94 88 | 15,90.21 89 | 15,96.91 90 | 15,93.12 91 | 15,97.13 92 | 15,96.94 93 | 15,95.39 94 | 15,95.33 95 | 15,96.57 96 | 15,95.43 97 | 15,96.30 98 | 15,96.29 99 | 15,95.86 100 | 15,95.92 101 | 15,95.32 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/MNISTMult2_vae_k10_poisson_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-08_22:37 2 | 15,91.50 3 | 15,91.43 4 | 15,92.23 5 | 15,93.15 6 | 15,89.51 7 | 15,92.69 8 | 15,93.43 9 | 15,86.32 10 | 15,92.94 11 | 15,92.96 12 | 15,94.43 13 | 15,86.88 14 | 15,93.61 15 | 15,94.46 16 | 15,91.42 17 | 15,93.68 18 | 15,92.77 19 | 15,95.03 20 | 15,95.24 21 | 15,93.75 22 | 15,91.17 23 | 15,92.43 24 | 15,91.25 25 | 15,93.23 26 | 15,91.18 27 | 15,93.78 28 | 15,93.97 29 | 15,88.67 30 | 15,92.23 31 | 15,93.44 32 | 15,94.02 33 | 15,91.62 34 | 15,92.53 35 | 15,82.86 36 | 15,91.35 37 | 15,94.70 38 | 15,94.71 39 | 15,94.14 40 | 15,94.73 41 | 15,90.99 42 | 15,88.83 43 | 15,94.18 44 | 15,94.67 45 | 15,93.04 46 | 15,87.00 47 | 15,95.34 48 | 15,92.06 49 | 15,92.96 50 | 15,94.79 51 | 15,75.29 52 | 15,94.47 53 | 15,93.29 54 | 15,94.52 55 | 15,92.84 56 | 15,91.21 57 | 15,90.00 58 | 15,93.10 59 | 15,91.40 60 | 15,90.32 61 | 15,94.08 62 | 15,94.95 63 | 15,87.24 64 | 15,94.49 65 | 15,94.21 66 | 15,94.55 67 | 15,92.65 68 | 15,94.86 69 | 15,93.58 70 | 15,93.47 71 | 15,92.77 72 | 15,81.21 73 | 15,91.40 74 | 15,91.90 75 | 15,89.31 76 | 15,94.44 77 | 15,95.04 78 | 15,95.29 79 | 15,94.35 80 | 15,86.46 81 | 15,93.88 82 | 15,93.89 83 | 15,94.72 84 | 15,93.68 85 | 15,94.49 86 | 15,88.04 87 | 15,83.36 88 | 15,94.31 89 | 15,95.38 90 | 15,93.94 91 | 15,94.01 92 | 15,94.68 93 | 15,92.36 94 | 15,93.69 95 | 15,89.66 96 | 15,92.41 97 | 15,80.41 98 | 15,95.22 99 | 15,92.97 100 | 15,92.94 101 | 15,90.99 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/MNISTMult3_vae_k10_modularitymbo_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2021-03-22_23:03 2 | 20,93.31 3 | 20,97.11 4 | 20,95.44 5 | 20,96.57 6 | 20,95.95 7 | 20,95.38 8 | 20,96.26 9 | 20,95.49 10 | 20,95.51 11 | 20,93.78 12 | 20,96.34 13 | 20,92.84 14 | 20,95.93 15 | 20,91.57 16 | 20,96.91 17 | 20,96.92 18 | 20,95.69 19 | 20,96.08 20 | 20,96.91 21 | 20,95.39 22 | 20,91.01 23 | 20,97.12 24 | 20,95.37 25 | 20,95.96 26 | 20,96.91 27 | 20,88.10 28 | 20,95.35 29 | 20,97.11 30 | 20,95.46 31 | 20,95.44 32 | 20,96.09 33 | 20,96.91 34 | 20,91.46 35 | 20,95.97 36 | 20,96.90 37 | 20,91.59 38 | 20,96.90 39 | 20,96.91 40 | 20,91.47 41 | 20,96.08 42 | 20,95.48 43 | 20,96.96 44 | 20,95.17 45 | 20,95.93 46 | 20,96.35 47 | 20,95.41 48 | 20,95.99 49 | 20,96.91 50 | 20,96.15 51 | 20,95.42 52 | 20,96.08 53 | 20,95.37 54 | 20,95.50 55 | 20,96.38 56 | 20,96.55 57 | 20,93.31 58 | 20,93.06 59 | 20,95.97 60 | 20,95.93 61 | 20,92.46 62 | 20,96.29 63 | 20,96.92 64 | 20,92.94 65 | 20,92.05 66 | 20,95.42 67 | 20,95.93 68 | 20,93.58 69 | 20,92.16 70 | 20,93.66 71 | 20,96.08 72 | 20,91.43 73 | 20,95.31 74 | 20,96.31 75 | 20,95.33 76 | 20,95.64 77 | 20,95.97 78 | 20,96.07 79 | 20,86.82 80 | 20,95.95 81 | 20,95.97 82 | 20,96.03 83 | 20,93.37 84 | 20,91.07 85 | 20,93.42 86 | 20,96.38 87 | 20,92.91 88 | 20,95.54 89 | 20,96.06 90 | 20,95.97 91 | 20,96.31 92 | 20,96.32 93 | 20,96.91 94 | 20,96.93 95 | 20,96.90 96 | 20,96.38 97 | 20,96.91 98 | 20,96.39 99 | 20,95.47 100 | 20,95.68 101 | 20,95.93 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/MNISTMult3_vae_k10_poisson_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-08_22:44 2 | 20,94.55 3 | 20,95.36 4 | 20,94.33 5 | 20,94.49 6 | 20,95.07 7 | 20,90.43 8 | 20,95.55 9 | 20,92.05 10 | 20,94.54 11 | 20,94.62 12 | 20,94.74 13 | 20,90.83 14 | 20,95.36 15 | 20,92.83 16 | 20,90.95 17 | 20,94.05 18 | 20,94.91 19 | 20,92.80 20 | 20,90.07 21 | 20,94.90 22 | 20,94.96 23 | 20,95.16 24 | 20,94.42 25 | 20,94.39 26 | 20,84.40 27 | 20,92.08 28 | 20,94.73 29 | 20,93.24 30 | 20,94.06 31 | 20,92.74 32 | 20,95.53 33 | 20,92.80 34 | 20,92.94 35 | 20,94.07 36 | 20,94.50 37 | 20,94.32 38 | 20,93.50 39 | 20,95.54 40 | 20,94.56 41 | 20,96.17 42 | 20,93.38 43 | 20,95.05 44 | 20,91.74 45 | 20,94.67 46 | 20,94.93 47 | 20,93.61 48 | 20,94.44 49 | 20,93.92 50 | 20,92.75 51 | 20,93.78 52 | 20,94.88 53 | 20,92.91 54 | 20,94.04 55 | 20,94.97 56 | 20,94.86 57 | 20,82.44 58 | 20,95.60 59 | 20,95.90 60 | 20,94.67 61 | 20,90.27 62 | 20,87.07 63 | 20,95.26 64 | 20,95.13 65 | 20,93.04 66 | 20,92.46 67 | 20,94.46 68 | 20,94.50 69 | 20,93.75 70 | 20,94.07 71 | 20,93.96 72 | 20,91.90 73 | 20,94.90 74 | 20,95.06 75 | 20,93.93 76 | 20,94.24 77 | 20,93.71 78 | 20,95.46 79 | 20,89.70 80 | 20,88.83 81 | 20,94.94 82 | 20,92.35 83 | 20,92.23 84 | 20,94.24 85 | 20,96.28 86 | 20,94.78 87 | 20,94.79 88 | 20,95.24 89 | 20,94.21 90 | 20,91.80 91 | 20,94.18 92 | 20,91.23 93 | 20,95.78 94 | 20,95.42 95 | 20,95.25 96 | 20,91.03 97 | 20,93.96 98 | 20,94.55 99 | 20,95.46 100 | 20,94.52 101 | 20,94.54 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/MNISTMult4_vae_k10_modularitymbo_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2021-03-22_23:03 2 | 25,93.35 3 | 25,95.41 4 | 25,96.94 5 | 25,95.61 6 | 25,96.38 7 | 25,95.34 8 | 25,92.84 9 | 25,95.29 10 | 25,97.12 11 | 25,95.99 12 | 25,95.53 13 | 25,92.55 14 | 25,96.94 15 | 25,95.95 16 | 25,96.03 17 | 25,95.48 18 | 25,95.97 19 | 25,95.97 20 | 25,95.50 21 | 25,95.51 22 | 25,96.93 23 | 25,96.90 24 | 25,96.90 25 | 25,96.94 26 | 25,96.32 27 | 25,96.07 28 | 25,96.93 29 | 25,95.36 30 | 25,96.08 31 | 25,96.91 32 | 25,95.97 33 | 25,96.92 34 | 25,96.38 35 | 25,93.05 36 | 25,93.35 37 | 25,91.72 38 | 25,78.68 39 | 25,93.39 40 | 25,95.99 41 | 25,96.35 42 | 25,95.98 43 | 25,96.32 44 | 25,96.34 45 | 25,96.12 46 | 25,96.35 47 | 25,96.29 48 | 25,95.97 49 | 25,92.74 50 | 25,96.00 51 | 25,92.52 52 | 25,96.16 53 | 25,96.93 54 | 25,96.29 55 | 25,95.48 56 | 25,96.28 57 | 25,87.72 58 | 25,92.30 59 | 25,95.51 60 | 25,95.97 61 | 25,95.97 62 | 25,95.51 63 | 25,96.86 64 | 25,92.05 65 | 25,95.47 66 | 25,96.91 67 | 25,84.05 68 | 25,96.34 69 | 25,96.94 70 | 25,95.45 71 | 25,92.76 72 | 25,95.95 73 | 25,92.46 74 | 25,96.93 75 | 25,96.90 76 | 25,96.00 77 | 25,95.93 78 | 25,96.93 79 | 25,96.04 80 | 25,94.97 81 | 25,96.05 82 | 25,95.61 83 | 25,96.91 84 | 25,96.36 85 | 25,96.91 86 | 25,95.37 87 | 25,95.97 88 | 25,96.06 89 | 25,95.44 90 | 25,91.08 91 | 25,95.45 92 | 25,96.93 93 | 25,95.54 94 | 25,93.68 95 | 25,95.96 96 | 25,95.33 97 | 25,96.08 98 | 25,97.08 99 | 25,95.42 100 | 25,96.31 101 | 25,96.13 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/MNISTMult4_vae_k10_poisson_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-08_22:52 2 | 25,95.12 3 | 25,95.00 4 | 25,94.06 5 | 25,94.88 6 | 25,95.70 7 | 25,93.59 8 | 25,95.32 9 | 25,93.92 10 | 25,95.18 11 | 25,91.54 12 | 25,94.89 13 | 25,80.83 14 | 25,94.90 15 | 25,92.44 16 | 25,92.29 17 | 25,95.25 18 | 25,95.64 19 | 25,95.34 20 | 25,94.51 21 | 25,92.74 22 | 25,93.90 23 | 25,91.83 24 | 25,93.59 25 | 25,93.84 26 | 25,95.05 27 | 25,91.74 28 | 25,95.27 29 | 25,94.26 30 | 25,93.57 31 | 25,95.64 32 | 25,92.87 33 | 25,94.24 34 | 25,94.61 35 | 25,95.37 36 | 25,93.45 37 | 25,93.30 38 | 25,94.80 39 | 25,82.86 40 | 25,95.83 41 | 25,94.79 42 | 25,95.23 43 | 25,94.76 44 | 25,93.70 45 | 25,94.30 46 | 25,92.99 47 | 25,95.30 48 | 25,94.93 49 | 25,95.68 50 | 25,91.79 51 | 25,93.74 52 | 25,93.99 53 | 25,94.68 54 | 25,94.24 55 | 25,95.80 56 | 25,91.91 57 | 25,92.52 58 | 25,91.19 59 | 25,94.86 60 | 25,94.81 61 | 25,94.11 62 | 25,93.69 63 | 25,93.93 64 | 25,94.18 65 | 25,93.03 66 | 25,94.81 67 | 25,86.13 68 | 25,91.61 69 | 25,95.65 70 | 25,94.59 71 | 25,93.50 72 | 25,96.00 73 | 25,93.73 74 | 25,94.24 75 | 25,94.21 76 | 25,92.18 77 | 25,89.76 78 | 25,93.37 79 | 25,92.34 80 | 25,92.89 81 | 25,94.70 82 | 25,95.09 83 | 25,95.25 84 | 25,94.86 85 | 25,93.78 86 | 25,91.42 87 | 25,93.99 88 | 25,95.28 89 | 25,95.13 90 | 25,94.24 91 | 25,94.81 92 | 25,93.55 93 | 25,95.24 94 | 25,92.27 95 | 25,94.32 96 | 25,93.39 97 | 25,93.77 98 | 25,94.58 99 | 25,95.39 100 | 25,94.50 101 | 25,94.91 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/MNISTMult5_vae_k10_modularitymbo_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2021-03-22_23:04 2 | 30,93.11 3 | 30,96.90 4 | 30,96.93 5 | 30,96.91 6 | 30,95.39 7 | 30,96.35 8 | 30,95.95 9 | 30,95.32 10 | 30,95.38 11 | 30,91.91 12 | 30,95.48 13 | 30,93.40 14 | 30,90.05 15 | 30,92.51 16 | 30,96.71 17 | 30,95.95 18 | 30,95.54 19 | 30,93.35 20 | 30,95.99 21 | 30,96.93 22 | 30,95.42 23 | 30,95.42 24 | 30,95.65 25 | 30,95.97 26 | 30,96.29 27 | 30,92.25 28 | 30,92.67 29 | 30,96.38 30 | 30,95.42 31 | 30,96.07 32 | 30,95.44 33 | 30,95.96 34 | 30,95.93 35 | 30,96.38 36 | 30,95.34 37 | 30,96.31 38 | 30,92.84 39 | 30,96.41 40 | 30,95.88 41 | 30,96.32 42 | 30,96.28 43 | 30,96.01 44 | 30,90.73 45 | 30,93.52 46 | 30,96.94 47 | 30,93.07 48 | 30,92.98 49 | 30,92.39 50 | 30,92.02 51 | 30,95.46 52 | 30,93.34 53 | 30,96.78 54 | 30,92.88 55 | 30,95.44 56 | 30,96.27 57 | 30,93.08 58 | 30,93.02 59 | 30,92.83 60 | 30,96.35 61 | 30,96.32 62 | 30,95.37 63 | 30,93.49 64 | 30,95.40 65 | 30,96.91 66 | 30,95.95 67 | 30,96.04 68 | 30,95.43 69 | 30,96.29 70 | 30,95.94 71 | 30,91.77 72 | 30,95.34 73 | 30,95.52 74 | 30,96.92 75 | 30,93.04 76 | 30,96.32 77 | 30,95.42 78 | 30,96.28 79 | 30,96.93 80 | 30,95.49 81 | 30,96.93 82 | 30,92.37 83 | 30,93.03 84 | 30,96.96 85 | 30,83.90 86 | 30,96.93 87 | 30,96.32 88 | 30,96.24 89 | 30,93.10 90 | 30,95.97 91 | 30,96.35 92 | 30,96.90 93 | 30,96.91 94 | 30,95.15 95 | 30,96.91 96 | 30,95.95 97 | 30,96.32 98 | 30,91.73 99 | 30,96.37 100 | 30,96.32 101 | 30,96.41 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/MNISTMult5_vae_k10_poisson_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-08_22:59 2 | 30,95.51 3 | 30,92.51 4 | 30,94.00 5 | 30,94.31 6 | 30,95.50 7 | 30,95.64 8 | 30,95.95 9 | 30,95.23 10 | 30,95.66 11 | 30,95.37 12 | 30,95.09 13 | 30,94.69 14 | 30,95.04 15 | 30,86.61 16 | 30,93.17 17 | 30,93.49 18 | 30,94.32 19 | 30,95.62 20 | 30,91.11 21 | 30,90.89 22 | 30,95.01 23 | 30,95.53 24 | 30,94.04 25 | 30,94.93 26 | 30,93.93 27 | 30,94.52 28 | 30,92.34 29 | 30,88.51 30 | 30,94.59 31 | 30,94.18 32 | 30,95.28 33 | 30,93.91 34 | 30,95.22 35 | 30,95.35 36 | 30,91.13 37 | 30,93.77 38 | 30,95.70 39 | 30,83.72 40 | 30,95.61 41 | 30,93.59 42 | 30,94.92 43 | 30,93.38 44 | 30,95.79 45 | 30,94.46 46 | 30,91.33 47 | 30,92.56 48 | 30,95.02 49 | 30,92.61 50 | 30,95.91 51 | 30,95.23 52 | 30,94.89 53 | 30,95.97 54 | 30,93.34 55 | 30,94.13 56 | 30,94.31 57 | 30,94.21 58 | 30,94.96 59 | 30,94.41 60 | 30,85.95 61 | 30,96.04 62 | 30,95.50 63 | 30,89.43 64 | 30,93.79 65 | 30,94.61 66 | 30,94.44 67 | 30,95.73 68 | 30,92.24 69 | 30,91.11 70 | 30,94.96 71 | 30,95.52 72 | 30,95.62 73 | 30,95.29 74 | 30,89.27 75 | 30,94.93 76 | 30,95.02 77 | 30,92.39 78 | 30,91.70 79 | 30,95.38 80 | 30,95.03 81 | 30,91.79 82 | 30,95.26 83 | 30,94.79 84 | 30,94.53 85 | 30,94.20 86 | 30,89.16 87 | 30,95.17 88 | 30,94.74 89 | 30,94.84 90 | 30,93.84 91 | 30,95.39 92 | 30,94.71 93 | 30,83.97 94 | 30,93.77 95 | 30,94.82 96 | 30,94.17 97 | 30,93.31 98 | 30,95.70 99 | 30,94.85 100 | 30,94.90 101 | 30,95.24 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/MNIST_vae_k10_laplace_accuracy.csv: -------------------------------------------------------------------------------- 1 | 40,53.51 2 | 30,17.93 3 | 20,37.86 4 | 10,11.46 5 | 50,69.00 6 | 20,45.98 7 | 10,12.66 8 | 30,31.71 9 | 40,59.93 10 | 50,71.08 11 | 10,11.51 12 | 20,34.39 13 | 30,37.32 14 | 50,56.47 15 | 40,50.07 16 | 10,21.00 17 | 20,41.28 18 | 30,33.04 19 | 40,63.25 20 | 50,64.57 21 | 10,20.14 22 | 20,27.52 23 | 40,39.76 24 | 30,33.67 25 | 50,42.80 26 | 10,29.53 27 | 20,24.77 28 | 30,42.38 29 | 40,57.95 30 | 50,64.64 31 | 20,35.17 32 | 10,10.15 33 | 30,54.32 34 | 40,72.51 35 | 50,58.97 36 | 10,12.29 37 | 20,38.88 38 | 30,68.21 39 | 40,68.52 40 | 50,90.58 41 | 10,11.32 42 | 20,20.88 43 | 30,33.92 44 | 40,44.23 45 | 50,56.91 46 | 10,10.10 47 | 30,39.36 48 | 20,30.70 49 | 40,59.24 50 | 50,63.49 51 | 10,10.65 52 | 20,12.89 53 | 30,25.56 54 | 40,38.47 55 | 50,64.78 56 | 10,10.35 57 | 20,21.80 58 | 30,26.20 59 | 50,65.03 60 | 40,57.67 61 | 10,10.12 62 | 20,39.10 63 | 30,41.90 64 | 40,71.27 65 | 50,64.16 66 | 10,10.86 67 | 20,26.48 68 | 30,37.44 69 | 40,74.68 70 | 50,87.98 71 | 10,20.44 72 | 20,20.89 73 | 30,40.89 74 | 40,57.18 75 | 50,56.10 76 | 10,13.81 77 | 20,14.60 78 | 30,25.44 79 | 40,36.02 80 | 50,77.85 81 | 10,11.30 82 | 20,44.80 83 | 30,48.91 84 | 40,80.03 85 | 50,44.63 86 | 10,12.85 87 | 20,15.29 88 | 30,42.64 89 | 40,60.31 90 | 50,75.74 91 | 10,10.65 92 | 20,25.02 93 | 30,28.00 94 | 40,50.39 95 | 50,90.19 96 | 10,11.61 97 | 20,17.57 98 | 30,51.08 99 | 40,64.17 100 | 50,52.16 101 | 10,11.08 102 | 20,14.46 103 | 30,34.11 104 | 40,72.56 105 | 50,60.75 106 | 10,10.20 107 | 20,31.72 108 | 30,30.93 109 | 40,54.18 110 | 50,68.83 111 | 10,19.36 112 | 20,42.15 113 | 30,49.64 114 | 40,41.55 115 | 50,32.42 116 | 10,9.32 117 | 20,10.90 118 | 30,47.44 119 | 40,62.54 120 | 50,65.79 121 | 10,11.06 122 | 20,46.88 123 | 30,50.89 124 | 40,54.40 125 | 50,74.16 126 | 20,21.56 127 | 10,17.59 128 | 30,63.05 129 | 40,53.94 130 | 50,76.66 131 | 10,20.99 132 | 20,14.72 133 | 30,59.47 134 | 40,70.88 135 | 50,84.04 136 | 10,10.78 137 | 20,31.80 138 | 30,46.56 139 | 40,50.87 140 | 50,80.07 141 | 10,13.84 142 | 40,61.97 143 | 20,21.81 144 | 30,28.45 145 | 50,75.79 146 | 30,32.92 147 | 10,21.48 148 | 20,40.78 149 | 40,84.88 150 | 50,64.29 151 | 20,38.95 152 | 10,10.44 153 | 30,31.59 154 | 50,63.01 155 | 40,54.44 156 | 10,24.40 157 | 20,30.98 158 | 40,66.29 159 | 30,51.70 160 | 50,65.96 161 | 10,21.58 162 | 30,50.16 163 | 20,21.36 164 | 40,59.36 165 | 50,75.18 166 | 20,42.60 167 | 10,10.38 168 | 30,31.07 169 | 40,74.36 170 | 50,90.35 171 | 20,28.30 172 | 10,24.72 173 | 30,37.89 174 | 40,68.39 175 | 50,76.18 176 | 10,10.85 177 | 20,30.96 178 | 30,35.69 179 | 40,68.54 180 | 50,65.97 181 | 10,10.84 182 | 20,31.75 183 | 40,32.05 184 | 30,62.53 185 | 50,56.89 186 | 30,49.70 187 | 20,42.98 188 | 10,21.55 189 | 40,40.69 190 | 50,70.68 191 | 10,10.00 192 | 20,33.09 193 | 30,21.97 194 | 40,46.00 195 | 50,50.88 196 | 10,24.42 197 | 20,16.18 198 | 30,53.97 199 | 50,56.11 200 | 40,54.16 201 | 10,20.67 202 | 20,30.87 203 | 40,45.17 204 | 30,39.99 205 | 50,79.88 206 | 10,26.13 207 | 30,39.98 208 | 20,19.81 209 | 40,56.29 210 | 50,84.43 211 | 10,10.98 212 | 20,18.98 213 | 30,31.49 214 | 40,81.32 215 | 50,60.53 216 | 10,20.56 217 | 20,19.32 218 | 30,41.05 219 | 40,42.82 220 | 50,75.09 221 | 10,24.97 222 | 20,10.79 223 | 30,52.63 224 | 40,48.67 225 | 50,77.95 226 | 10,12.02 227 | 20,29.08 228 | 30,46.37 229 | 40,88.66 230 | 50,48.49 231 | 20,30.31 232 | 10,17.59 233 | 30,60.26 234 | 40,46.45 235 | 50,71.32 236 | 10,12.41 237 | 20,15.62 238 | 30,32.10 239 | 40,45.70 240 | 50,68.62 241 | 10,21.19 242 | 20,40.84 243 | 30,49.89 244 | 40,62.59 245 | 50,78.82 246 | 10,21.27 247 | 20,13.21 248 | 30,42.05 249 | 40,62.13 250 | 50,88.05 251 | 10,20.37 252 | 20,41.14 253 | 30,31.25 254 | 40,35.34 255 | 50,80.48 256 | 10,10.70 257 | 20,46.53 258 | 30,18.96 259 | 40,55.66 260 | 50,79.40 261 | 10,18.93 262 | 20,29.11 263 | 30,41.78 264 | 40,71.94 265 | 50,66.56 266 | 10,10.90 267 | 20,26.18 268 | 30,68.55 269 | 40,40.07 270 | 50,71.09 271 | 10,16.32 272 | 20,42.80 273 | 30,26.10 274 | 40,35.03 275 | 50,70.32 276 | 10,23.01 277 | 20,36.68 278 | 30,44.24 279 | 40,61.60 280 | 50,73.78 281 | 10,23.17 282 | 20,13.95 283 | 30,49.97 284 | 40,73.26 285 | 50,66.01 286 | 10,12.17 287 | 20,22.45 288 | 30,55.80 289 | 40,61.52 290 | 50,71.79 291 | 10,20.91 292 | 20,11.89 293 | 30,43.14 294 | 40,59.07 295 | 50,90.43 296 | 10,10.54 297 | 20,26.06 298 | 30,34.43 299 | 40,63.76 300 | 50,80.11 301 | 10,19.27 302 | 20,31.09 303 | 30,51.45 304 | 40,42.86 305 | 50,83.95 306 | 10,35.03 307 | 20,33.18 308 | 30,38.64 309 | 40,52.96 310 | 50,70.72 311 | 10,11.41 312 | 20,23.01 313 | 30,50.74 314 | 40,77.36 315 | 50,75.58 316 | 10,10.38 317 | 20,25.16 318 | 30,23.45 319 | 40,69.99 320 | 50,70.78 321 | 10,13.83 322 | 20,21.15 323 | 30,56.34 324 | 40,63.76 325 | 50,76.07 326 | 10,10.45 327 | 20,33.95 328 | 30,44.76 329 | 40,57.28 330 | 50,74.51 331 | 10,11.63 332 | 20,26.57 333 | 30,41.14 334 | 40,40.77 335 | 50,77.17 336 | 10,12.50 337 | 20,24.82 338 | 30,63.01 339 | 40,52.82 340 | 50,75.25 341 | 10,19.87 342 | 20,14.36 343 | 30,58.58 344 | 40,38.11 345 | 50,87.04 346 | 10,10.86 347 | 30,24.68 348 | 20,21.41 349 | 40,51.29 350 | 50,64.34 351 | 20,40.53 352 | 10,12.68 353 | 30,42.46 354 | 40,74.72 355 | 50,41.61 356 | 10,14.00 357 | 20,16.96 358 | 30,51.52 359 | 40,40.81 360 | 50,62.06 361 | 10,23.71 362 | 20,26.53 363 | 30,31.97 364 | 40,48.69 365 | 50,68.73 366 | 10,21.92 367 | 20,30.02 368 | 30,53.31 369 | 40,48.92 370 | 50,64.19 371 | 10,38.37 372 | 20,22.84 373 | 30,32.95 374 | 40,62.46 375 | 50,86.33 376 | 10,14.62 377 | 20,57.47 378 | 30,33.98 379 | 40,61.98 380 | 50,66.35 381 | 10,22.65 382 | 20,45.13 383 | 30,22.20 384 | 40,42.20 385 | 50,69.99 386 | 10,10.35 387 | 20,28.85 388 | 30,38.32 389 | 40,59.31 390 | 50,69.34 391 | 10,10.90 392 | 20,26.32 393 | 30,48.65 394 | 40,57.39 395 | 50,87.34 396 | 10,22.18 397 | 20,20.57 398 | 30,41.06 399 | 40,69.60 400 | 50,59.43 401 | 10,21.82 402 | 20,19.00 403 | 30,55.82 404 | 40,53.54 405 | 50,64.97 406 | 20,22.46 407 | 10,29.16 408 | 30,40.56 409 | 40,51.82 410 | 50,69.05 411 | 10,12.24 412 | 20,31.07 413 | 30,39.64 414 | 40,53.13 415 | 50,38.80 416 | 10,12.08 417 | 20,24.44 418 | 30,43.84 419 | 40,35.97 420 | 50,79.37 421 | 10,10.08 422 | 20,33.56 423 | 30,59.60 424 | 40,61.15 425 | 50,77.19 426 | 10,11.45 427 | 20,16.31 428 | 30,37.15 429 | 40,62.31 430 | 50,55.12 431 | 10,27.49 432 | 20,20.57 433 | 30,54.06 434 | 40,48.84 435 | 50,63.10 436 | 10,9.44 437 | 20,17.93 438 | 30,45.09 439 | 40,55.72 440 | 50,72.48 441 | 10,21.98 442 | 20,11.63 443 | 30,38.30 444 | 40,54.30 445 | 50,74.85 446 | 10,14.40 447 | 20,42.70 448 | 30,42.08 449 | 40,73.80 450 | 50,64.75 451 | 10,15.55 452 | 20,25.44 453 | 30,40.20 454 | 40,79.53 455 | 50,64.89 456 | 10,12.60 457 | 20,15.89 458 | 30,29.60 459 | 40,60.25 460 | 50,85.26 461 | 10,10.12 462 | 20,31.42 463 | 30,89.86 464 | 40,57.99 465 | 50,83.63 466 | 10,11.32 467 | 20,41.66 468 | 30,24.13 469 | 40,60.07 470 | 50,53.14 471 | 10,22.36 472 | 20,23.60 473 | 30,25.85 474 | 40,59.96 475 | 50,93.12 476 | 10,22.85 477 | 20,39.65 478 | 30,31.21 479 | 40,66.86 480 | 50,63.03 481 | 10,12.49 482 | 20,40.95 483 | 30,43.13 484 | 40,75.50 485 | 50,79.61 486 | 10,15.88 487 | 20,32.88 488 | 30,46.60 489 | 40,50.89 490 | 50,70.37 491 | 10,10.42 492 | 20,43.83 493 | 30,28.14 494 | 40,75.23 495 | 50,53.36 496 | 10,13.11 497 | 20,19.47 498 | 30,53.30 499 | 40,62.42 500 | 50,54.97 501 | -------------------------------------------------------------------------------- /ResultsFromPaper/MNIST_vae_k10_mbo_accuracy.csv: -------------------------------------------------------------------------------- 1 | 10,23.36 2 | 30,40.39 3 | 20,28.91 4 | 40,51.03 5 | 50,56.49 6 | 10,19.60 7 | 20,29.06 8 | 30,45.63 9 | 40,57.89 10 | 10,25.97 11 | 50,59.85 12 | 20,40.77 13 | 30,36.14 14 | 40,53.01 15 | 50,48.98 16 | 10,18.23 17 | 20,21.60 18 | 30,40.44 19 | 40,51.19 20 | 50,68.01 21 | 10,17.74 22 | 20,34.24 23 | 30,30.73 24 | 40,52.49 25 | 50,56.60 26 | 10,23.55 27 | 20,24.21 28 | 30,48.12 29 | 40,46.36 30 | 50,62.34 31 | 10,11.24 32 | 20,32.23 33 | 30,46.58 34 | 40,59.12 35 | 50,58.15 36 | 10,16.04 37 | 20,21.98 38 | 30,36.79 39 | 40,46.73 40 | 50,52.15 41 | 10,17.19 42 | 20,20.67 43 | 30,41.03 44 | 40,43.29 45 | 50,60.61 46 | 10,18.83 47 | 20,27.03 48 | 30,49.06 49 | 40,49.79 50 | 50,60.14 51 | 10,25.83 52 | 20,19.02 53 | 30,30.90 54 | 40,55.51 55 | 50,54.65 56 | 10,24.54 57 | 20,19.40 58 | 30,50.02 59 | 40,50.98 60 | 50,61.47 61 | 10,26.39 62 | 20,20.17 63 | 30,32.12 64 | 40,61.38 65 | 50,52.03 66 | 10,20.70 67 | 20,42.91 68 | 30,28.16 69 | 40,48.04 70 | 50,64.91 71 | 10,25.26 72 | 20,26.74 73 | 30,40.33 74 | 40,47.13 75 | 50,55.24 76 | 10,13.54 77 | 20,30.40 78 | 30,32.57 79 | 40,56.05 80 | 50,64.48 81 | 10,19.12 82 | 20,30.53 83 | 30,39.95 84 | 40,49.52 85 | 50,65.95 86 | 10,21.23 87 | 20,26.88 88 | 30,49.30 89 | 40,54.57 90 | 50,51.25 91 | 10,27.20 92 | 20,20.16 93 | 30,54.03 94 | 40,59.17 95 | 50,55.82 96 | 10,12.04 97 | 20,24.47 98 | 30,35.55 99 | 40,52.92 100 | 50,71.28 101 | 10,23.29 102 | 20,25.30 103 | 30,50.93 104 | 40,47.29 105 | 50,64.91 106 | 10,16.96 107 | 20,30.32 108 | 30,26.46 109 | 40,49.56 110 | 50,56.75 111 | 10,19.94 112 | 20,24.05 113 | 30,47.31 114 | 50,57.80 115 | 40,41.83 116 | 10,38.97 117 | 20,24.44 118 | 40,52.28 119 | 30,44.24 120 | 50,59.29 121 | 10,29.52 122 | 20,21.24 123 | 30,50.57 124 | 40,44.75 125 | 50,65.38 126 | 10,12.12 127 | 20,36.06 128 | 30,26.65 129 | 40,43.39 130 | 50,52.13 131 | 10,26.14 132 | 20,36.00 133 | 30,39.87 134 | 40,47.76 135 | 50,62.81 136 | 10,19.31 137 | 20,32.28 138 | 30,33.55 139 | 40,57.05 140 | 50,58.79 141 | 10,15.68 142 | 20,23.10 143 | 30,34.34 144 | 40,38.29 145 | 50,46.01 146 | 10,16.84 147 | 20,37.10 148 | 30,48.99 149 | 40,51.59 150 | 50,58.00 151 | 10,11.90 152 | 20,23.26 153 | 30,40.92 154 | 40,47.46 155 | 50,62.02 156 | 10,14.05 157 | 20,28.34 158 | 30,39.12 159 | 40,47.10 160 | 50,53.99 161 | 10,14.21 162 | 20,32.94 163 | 30,48.32 164 | 40,49.87 165 | 50,61.17 166 | 10,11.49 167 | 20,27.45 168 | 30,44.56 169 | 40,57.31 170 | 50,45.03 171 | 10,14.49 172 | 20,28.59 173 | 30,49.14 174 | 40,51.34 175 | 50,58.77 176 | 10,21.89 177 | 20,38.83 178 | 30,36.57 179 | 40,53.09 180 | 50,59.15 181 | 10,15.89 182 | 20,23.02 183 | 30,45.24 184 | 40,49.71 185 | 50,53.93 186 | 10,17.11 187 | 20,27.82 188 | 30,32.57 189 | 40,55.79 190 | 50,54.63 191 | 10,16.57 192 | 20,30.42 193 | 30,44.38 194 | 40,63.67 195 | 50,62.91 196 | 10,16.58 197 | 20,34.59 198 | 30,36.76 199 | 40,52.29 200 | 50,62.11 201 | 20,25.31 202 | 10,21.77 203 | 30,36.98 204 | 40,47.80 205 | 50,65.75 206 | 10,21.18 207 | 20,21.03 208 | 30,37.27 209 | 40,48.80 210 | 50,43.00 211 | 10,22.32 212 | 20,29.61 213 | 30,33.21 214 | 40,53.54 215 | 50,61.57 216 | 10,16.19 217 | 20,44.33 218 | 30,48.70 219 | 40,45.16 220 | 50,57.59 221 | 10,32.73 222 | 20,23.90 223 | 30,35.09 224 | 40,44.92 225 | 50,52.00 226 | 10,9.31 227 | 20,24.61 228 | 30,55.97 229 | 40,56.65 230 | 50,63.19 231 | 10,21.34 232 | 20,25.00 233 | 30,35.65 234 | 40,45.36 235 | 50,57.49 236 | 10,22.73 237 | 30,38.78 238 | 20,24.59 239 | 40,49.28 240 | 50,56.61 241 | 10,24.01 242 | 20,24.33 243 | 30,45.41 244 | 40,44.70 245 | 50,56.42 246 | 10,15.27 247 | 20,35.07 248 | 30,42.09 249 | 40,51.03 250 | 50,58.65 251 | 10,19.12 252 | 20,33.29 253 | 30,41.61 254 | 40,40.21 255 | 50,71.74 256 | 10,20.05 257 | 20,33.55 258 | 30,35.20 259 | 40,48.76 260 | 50,72.83 261 | 10,14.82 262 | 20,39.86 263 | 30,46.51 264 | 40,28.91 265 | 50,56.23 266 | 10,28.29 267 | 20,32.65 268 | 30,48.80 269 | 40,47.88 270 | 50,53.95 271 | 10,17.78 272 | 20,21.53 273 | 30,37.34 274 | 40,44.45 275 | 50,60.11 276 | 10,15.30 277 | 20,31.40 278 | 30,32.61 279 | 40,54.50 280 | 50,58.82 281 | 10,14.24 282 | 20,26.52 283 | 30,46.88 284 | 40,44.93 285 | 50,56.27 286 | 10,17.94 287 | 20,26.30 288 | 30,39.14 289 | 40,58.24 290 | 50,57.30 291 | 10,19.72 292 | 20,30.64 293 | 30,37.43 294 | 40,51.51 295 | 50,63.59 296 | 10,30.98 297 | 20,30.16 298 | 30,41.69 299 | 40,57.50 300 | 50,66.70 301 | 10,16.80 302 | 20,29.79 303 | 30,53.92 304 | 40,51.97 305 | 50,72.82 306 | 10,24.15 307 | 20,39.54 308 | 30,53.42 309 | 40,49.18 310 | 50,54.39 311 | 10,24.97 312 | 20,38.85 313 | 30,48.30 314 | 40,41.57 315 | 50,47.94 316 | 10,11.56 317 | 20,32.15 318 | 30,31.44 319 | 40,53.51 320 | 50,65.91 321 | 10,15.46 322 | 20,41.11 323 | 30,51.76 324 | 40,51.24 325 | 50,68.44 326 | 10,30.41 327 | 20,33.82 328 | 30,35.19 329 | 40,52.17 330 | 50,63.80 331 | 10,27.45 332 | 20,30.28 333 | 30,32.93 334 | 40,48.85 335 | 50,61.40 336 | 10,26.99 337 | 20,33.22 338 | 30,33.01 339 | 40,50.31 340 | 50,56.87 341 | 10,29.18 342 | 20,20.92 343 | 30,31.44 344 | 40,40.09 345 | 50,58.58 346 | 10,14.61 347 | 20,17.24 348 | 30,33.04 349 | 40,53.02 350 | 50,57.21 351 | 10,10.83 352 | 20,19.81 353 | 30,29.27 354 | 40,64.41 355 | 50,61.35 356 | 10,11.59 357 | 20,31.66 358 | 30,52.27 359 | 40,55.40 360 | 50,54.59 361 | 10,17.66 362 | 20,40.62 363 | 30,37.67 364 | 40,38.38 365 | 50,54.60 366 | 10,10.14 367 | 20,31.36 368 | 30,42.17 369 | 40,52.98 370 | 50,59.21 371 | 10,19.00 372 | 20,26.26 373 | 30,34.18 374 | 40,41.99 375 | 50,56.90 376 | 10,18.10 377 | 20,29.98 378 | 30,33.85 379 | 40,49.08 380 | 50,55.30 381 | 10,14.47 382 | 20,38.41 383 | 30,20.14 384 | 40,55.09 385 | 50,62.69 386 | 10,23.91 387 | 20,31.35 388 | 30,38.55 389 | 40,52.03 390 | 50,63.66 391 | 10,4.77 392 | 20,26.60 393 | 30,42.20 394 | 40,54.92 395 | 50,61.65 396 | 10,8.26 397 | 20,38.39 398 | 30,42.92 399 | 40,59.26 400 | 10,25.29 401 | 20,28.12 402 | 30,47.32 403 | 40,57.94 404 | 50,50.54 405 | 20,31.59 406 | 10,21.71 407 | 30,21.93 408 | 40,63.75 409 | 50,65.47 410 | 10,15.72 411 | 20,39.51 412 | 30,26.47 413 | 40,54.29 414 | 50,62.27 415 | 10,28.43 416 | 20,17.00 417 | 30,45.55 418 | 40,47.38 419 | 50,57.58 420 | 10,29.54 421 | 20,38.74 422 | 30,49.62 423 | 40,56.31 424 | 50,57.66 425 | 10,19.21 426 | 20,27.66 427 | 30,43.80 428 | 40,51.09 429 | 50,63.84 430 | 10,15.69 431 | 20,36.39 432 | 30,50.95 433 | 40,49.19 434 | 50,65.70 435 | 10,27.71 436 | 20,32.37 437 | 30,33.21 438 | 40,53.92 439 | 50,65.50 440 | 10,15.87 441 | 20,52.14 442 | 30,43.23 443 | 40,48.91 444 | 50,63.76 445 | 10,22.42 446 | 20,29.39 447 | 30,36.19 448 | 40,43.48 449 | 50,64.83 450 | 10,29.49 451 | 20,34.31 452 | 30,39.33 453 | 40,54.41 454 | 50,44.46 455 | 10,12.13 456 | 20,19.95 457 | 30,37.60 458 | 40,40.57 459 | 50,56.53 460 | 10,16.45 461 | 20,20.12 462 | 30,40.12 463 | 40,43.04 464 | 50,59.24 465 | 10,18.64 466 | 30,38.53 467 | 40,54.96 468 | 50,52.97 469 | 10,14.30 470 | 20,17.15 471 | 30,40.25 472 | 40,45.96 473 | 50,62.83 474 | 10,14.49 475 | 20,24.35 476 | 30,41.05 477 | 40,61.16 478 | 50,67.58 479 | 10,17.32 480 | 20,31.86 481 | 30,45.49 482 | 40,57.83 483 | 50,54.42 484 | 10,19.29 485 | 20,23.07 486 | 30,40.38 487 | 40,49.25 488 | 50,59.71 489 | 10,22.12 490 | 20,20.90 491 | 30,40.67 492 | 40,51.23 493 | 50,57.44 494 | 10,6.09 495 | 20,32.66 496 | 30,38.11 497 | 40,55.32 498 | 50,68.69 499 | -------------------------------------------------------------------------------- /ResultsFromPaper/MNIST_vae_k10_p10laplace_none_accuracy.csv: -------------------------------------------------------------------------------- 1 | 50,91.20 2 | 40,92.46 3 | 30,89.16 4 | 20,86.94 5 | 10,69.73 6 | 10,61.24 7 | 30,89.93 8 | 20,91.43 9 | 40,92.32 10 | 50,91.09 11 | 20,89.22 12 | 40,90.37 13 | 50,90.64 14 | 30,91.85 15 | 10,77.04 16 | 10,86.65 17 | 50,94.07 18 | 30,85.74 19 | 40,91.50 20 | 20,83.86 21 | 10,77.36 22 | 50,91.11 23 | 40,88.11 24 | 30,88.68 25 | 20,89.70 26 | 10,81.22 27 | 50,92.97 28 | 40,88.72 29 | 20,80.96 30 | 30,89.62 31 | 10,62.32 32 | 20,92.19 33 | 40,90.62 34 | 50,91.62 35 | 30,90.49 36 | 10,65.66 37 | 50,92.61 38 | 40,87.54 39 | 30,90.69 40 | 20,82.04 41 | 10,81.63 42 | 20,81.80 43 | 50,91.74 44 | 30,91.57 45 | 40,90.52 46 | 20,87.20 47 | 30,89.56 48 | 40,90.89 49 | 50,91.56 50 | 10,60.44 51 | -------------------------------------------------------------------------------- /ResultsFromPaper/MNIST_vae_k10_poisson_accuracy50trials.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-08_14:35 2 | 20,94.48 3 | 20,94.62 4 | 50,95.96 5 | 10,91.51 6 | 10,92.92 7 | 40,95.73 8 | 40,95.40 9 | 30,94.79 10 | 20,92.48 11 | 10,91.12 12 | 50,94.99 13 | 30,95.15 14 | 50,96.12 15 | 40,95.28 16 | 40,95.19 17 | 30,93.05 18 | 10,94.80 19 | 50,95.81 20 | 30,94.02 21 | 20,92.04 22 | 10,93.04 23 | 30,95.10 24 | 20,95.62 25 | 50,95.68 26 | 40,93.45 27 | 20,94.14 28 | 10,91.97 29 | 30,95.04 30 | 20,95.77 31 | 50,96.28 32 | 10,81.75 33 | 40,95.34 34 | 40,94.87 35 | 30,95.46 36 | 10,91.80 37 | 10,92.62 38 | 50,95.73 39 | 30,94.82 40 | 40,94.73 41 | 50,95.29 42 | 20,91.36 43 | 50,95.42 44 | 20,95.00 45 | 30,95.50 46 | 40,95.14 47 | 10,85.52 48 | 20,94.99 49 | 30,95.28 50 | 50,94.37 51 | 40,95.56 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/MNIST_vae_k10_sparselabelpropagation_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-02_13:05 2 | 40,5.94 3 | 30,20.31 4 | 50,12.85 5 | 10,5.45 6 | 50,20.72 7 | 20,14.39 8 | 40,22.09 9 | 20,9.94 10 | 30,12.28 11 | 10,20.20 12 | 20,17.83 13 | 40,27.81 14 | 10,11.68 15 | 40,18.14 16 | 50,10.82 17 | 30,19.49 18 | 20,9.63 19 | 30,14.58 20 | 10,21.64 21 | 50,14.81 22 | 30,14.11 23 | 20,8.84 24 | 10,21.35 25 | 20,15.63 26 | 50,13.70 27 | 30,15.94 28 | 40,23.94 29 | 40,17.82 30 | 10,16.45 31 | 50,14.46 32 | 10,8.24 33 | 30,6.01 34 | 20,20.76 35 | 10,15.18 36 | 20,13.45 37 | 30,13.80 38 | 40,20.26 39 | 50,22.77 40 | 50,18.14 41 | 40,12.88 42 | 10,10.16 43 | 20,18.71 44 | 30,11.12 45 | 40,12.71 46 | 50,11.38 47 | 10,9.77 48 | 30,17.00 49 | 20,10.45 50 | 40,17.91 51 | 50,21.97 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/MNIST_vae_k10_wnll_accuracy.csv: -------------------------------------------------------------------------------- 1 | 40,95.81 2 | 20,89.02 3 | 30,88.35 4 | 10,75.98 5 | 50,93.96 6 | 20,77.09 7 | 10,70.73 8 | 30,93.30 9 | 40,95.43 10 | 50,93.79 11 | 10,65.70 12 | 20,92.56 13 | 30,92.81 14 | 40,94.73 15 | 50,96.00 16 | 10,76.97 17 | 20,72.93 18 | 30,88.64 19 | 40,94.97 20 | 50,96.03 21 | 10,63.11 22 | 20,86.78 23 | 40,93.82 24 | 30,91.38 25 | 50,95.35 26 | 30,93.96 27 | 10,45.45 28 | 20,82.87 29 | 40,92.28 30 | 50,94.36 31 | 20,83.37 32 | 10,58.46 33 | 40,92.64 34 | 30,93.17 35 | 50,95.71 36 | 10,54.78 37 | 30,93.04 38 | 20,86.43 39 | 40,92.36 40 | 50,94.36 41 | 10,39.32 42 | 20,70.29 43 | 40,95.05 44 | 30,91.64 45 | 50,93.82 46 | 10,81.11 47 | 20,57.97 48 | 30,92.87 49 | 40,94.82 50 | 50,91.94 51 | 10,56.49 52 | 20,70.94 53 | 30,92.23 54 | 40,91.61 55 | 50,94.91 56 | 10,17.53 57 | 20,94.88 58 | 30,85.60 59 | 40,94.69 60 | 50,94.81 61 | 10,63.35 62 | 20,70.23 63 | 30,93.27 64 | 40,92.53 65 | 50,93.73 66 | 10,53.65 67 | 20,88.59 68 | 30,83.11 69 | 40,94.17 70 | 50,94.52 71 | 10,37.20 72 | 20,84.58 73 | 30,90.31 74 | 40,94.78 75 | 50,95.72 76 | 10,50.97 77 | 20,87.86 78 | 30,91.63 79 | 40,95.38 80 | 50,94.53 81 | 10,30.36 82 | 20,82.82 83 | 30,82.79 84 | 40,93.36 85 | 50,95.82 86 | 10,77.71 87 | 20,83.59 88 | 30,85.97 89 | 40,93.55 90 | 50,94.84 91 | 10,76.36 92 | 20,76.30 93 | 40,92.29 94 | 30,81.81 95 | 50,94.55 96 | 10,63.97 97 | 30,92.88 98 | 20,78.51 99 | 40,88.73 100 | 50,93.89 101 | 10,46.37 102 | 20,90.76 103 | 30,88.53 104 | 40,95.10 105 | 50,94.13 106 | 10,59.74 107 | 20,89.51 108 | 30,88.76 109 | 40,94.76 110 | 50,94.46 111 | 10,47.52 112 | 20,92.83 113 | 30,94.62 114 | 40,93.15 115 | 50,94.22 116 | 10,53.19 117 | 30,92.58 118 | 20,73.47 119 | 40,93.05 120 | 50,93.87 121 | 10,69.70 122 | 20,65.00 123 | 30,92.90 124 | 40,91.28 125 | 50,94.90 126 | 10,53.06 127 | 20,84.45 128 | 30,90.47 129 | 40,92.79 130 | 50,95.25 131 | 10,52.46 132 | 20,71.98 133 | 30,89.72 134 | 40,95.24 135 | 50,95.74 136 | 10,47.08 137 | 20,76.05 138 | 30,90.33 139 | 40,95.22 140 | 50,94.79 141 | 10,60.36 142 | 20,87.24 143 | 30,89.41 144 | 40,94.95 145 | 50,95.28 146 | 10,80.26 147 | 20,87.78 148 | 30,89.68 149 | 40,94.71 150 | 50,93.08 151 | 10,24.21 152 | 20,85.36 153 | 30,90.81 154 | 40,94.50 155 | 50,95.76 156 | 10,38.91 157 | 20,78.15 158 | 30,89.03 159 | 40,93.75 160 | 50,93.06 161 | 10,74.58 162 | 20,86.52 163 | 30,80.56 164 | 40,94.67 165 | 50,93.63 166 | 10,39.49 167 | 20,69.71 168 | 30,93.46 169 | 40,95.21 170 | 50,95.82 171 | 10,67.00 172 | 20,81.39 173 | 30,94.79 174 | 40,94.07 175 | 50,95.69 176 | 20,86.51 177 | 10,65.57 178 | 30,89.30 179 | 40,95.25 180 | 50,94.85 181 | 10,72.27 182 | 20,93.85 183 | 30,87.67 184 | 40,96.05 185 | 50,93.89 186 | 10,65.76 187 | 20,88.97 188 | 30,83.53 189 | 40,92.30 190 | 50,94.55 191 | 10,84.02 192 | 20,81.07 193 | 30,91.33 194 | 40,95.41 195 | 50,93.93 196 | 10,50.61 197 | 20,82.30 198 | 30,93.65 199 | 40,93.83 200 | 50,96.12 201 | 10,46.81 202 | 20,92.66 203 | 30,90.08 204 | 40,92.09 205 | 50,94.30 206 | 10,53.03 207 | 20,76.90 208 | 30,94.43 209 | 40,95.53 210 | 50,94.96 211 | 10,79.30 212 | 30,89.69 213 | 20,88.27 214 | 40,93.22 215 | 50,94.86 216 | 20,76.98 217 | 10,58.89 218 | 30,91.57 219 | 40,95.13 220 | 50,93.66 221 | 10,71.41 222 | 20,93.95 223 | 30,89.91 224 | 40,94.34 225 | 50,95.56 226 | 10,67.69 227 | 20,91.18 228 | 40,95.12 229 | 30,94.00 230 | 50,91.73 231 | 10,58.00 232 | 20,84.51 233 | 40,93.74 234 | 30,95.42 235 | 50,96.19 236 | 30,87.92 237 | 20,92.72 238 | 10,44.74 239 | 40,94.55 240 | 50,94.70 241 | 10,45.40 242 | 20,81.98 243 | 30,88.22 244 | 40,91.48 245 | 50,95.01 246 | 10,46.19 247 | 20,69.39 248 | 30,91.90 249 | 40,93.65 250 | 50,95.84 251 | 10,41.27 252 | 20,74.10 253 | 30,87.63 254 | 40,92.12 255 | 50,96.30 256 | 20,90.43 257 | 10,57.47 258 | 30,81.52 259 | 40,93.46 260 | 50,94.67 261 | 10,41.85 262 | 20,81.27 263 | 30,93.18 264 | 40,93.69 265 | 50,94.22 266 | 10,35.64 267 | 20,78.52 268 | 30,94.16 269 | 40,91.33 270 | 50,93.23 271 | 10,61.11 272 | 20,72.57 273 | 30,87.66 274 | 40,92.15 275 | 50,95.19 276 | 10,73.61 277 | 20,87.51 278 | 30,86.50 279 | 40,93.20 280 | 50,95.71 281 | 10,57.86 282 | 20,85.29 283 | 30,94.64 284 | 40,95.54 285 | 50,91.84 286 | 10,56.32 287 | 20,89.46 288 | 30,92.17 289 | 40,94.39 290 | 50,95.70 291 | 10,47.34 292 | 20,91.42 293 | 30,91.89 294 | 40,89.68 295 | 50,95.11 296 | 10,73.52 297 | 20,82.59 298 | 30,93.47 299 | 40,94.43 300 | 50,95.38 301 | 10,72.98 302 | 20,76.35 303 | 30,87.10 304 | 40,94.69 305 | 50,95.44 306 | 10,63.15 307 | 20,84.68 308 | 30,95.86 309 | 40,94.24 310 | 50,95.66 311 | 10,51.09 312 | 20,82.91 313 | 30,89.30 314 | 40,94.59 315 | 50,94.77 316 | 10,38.35 317 | 20,73.31 318 | 30,94.55 319 | 40,91.82 320 | 50,94.99 321 | 10,64.85 322 | 20,87.40 323 | 30,92.29 324 | 40,92.49 325 | 50,94.50 326 | 10,66.03 327 | 20,88.50 328 | 30,91.82 329 | 40,95.83 330 | 50,95.36 331 | 10,78.72 332 | 20,80.16 333 | 30,94.14 334 | 40,90.24 335 | 50,94.52 336 | 10,37.99 337 | 20,88.83 338 | 30,91.54 339 | 40,92.79 340 | 50,96.51 341 | 10,56.89 342 | 20,65.25 343 | 30,92.67 344 | 40,90.74 345 | 50,94.85 346 | 10,35.39 347 | 20,63.45 348 | 40,94.53 349 | 30,88.44 350 | 50,95.21 351 | 10,73.74 352 | 30,92.86 353 | 20,82.86 354 | 40,93.43 355 | 50,95.38 356 | 20,88.82 357 | 10,57.46 358 | 30,84.36 359 | 40,94.46 360 | 10,59.31 361 | 50,93.07 362 | 20,85.53 363 | 30,92.35 364 | 50,93.48 365 | 40,95.17 366 | 10,39.12 367 | 20,85.75 368 | 40,92.19 369 | 30,93.78 370 | 50,90.08 371 | 10,59.12 372 | 20,89.30 373 | 30,90.15 374 | 40,95.02 375 | 50,93.38 376 | 10,54.82 377 | 20,82.24 378 | 40,94.65 379 | 30,87.40 380 | 50,95.27 381 | 10,72.55 382 | 30,89.81 383 | 20,90.09 384 | 40,94.88 385 | 50,95.57 386 | 10,31.13 387 | 20,80.35 388 | 30,93.81 389 | 40,92.88 390 | 50,94.89 391 | 10,62.12 392 | 20,93.83 393 | 30,91.10 394 | 40,93.73 395 | 50,95.31 396 | 10,43.64 397 | 20,82.05 398 | 30,90.62 399 | 40,93.12 400 | 50,94.20 401 | 10,23.56 402 | 20,77.57 403 | 30,88.41 404 | 40,93.34 405 | 50,95.46 406 | 10,54.87 407 | 20,89.24 408 | 30,87.06 409 | 40,93.31 410 | 50,95.38 411 | 10,81.54 412 | 20,82.88 413 | 30,86.82 414 | 40,91.31 415 | 50,94.18 416 | 10,57.01 417 | 20,78.27 418 | 30,93.29 419 | 40,95.85 420 | 50,94.42 421 | 10,41.37 422 | 20,84.03 423 | 30,85.66 424 | 40,94.63 425 | 50,95.71 426 | 10,63.84 427 | 20,89.09 428 | 30,94.09 429 | 40,94.59 430 | 50,95.64 431 | 20,87.34 432 | 10,41.44 433 | 30,92.11 434 | 40,94.27 435 | 50,94.76 436 | 10,62.76 437 | 20,82.05 438 | 30,93.13 439 | 40,93.61 440 | 50,95.01 441 | 10,70.18 442 | 20,84.83 443 | 30,91.19 444 | 40,93.48 445 | 50,94.71 446 | 10,22.04 447 | 20,89.93 448 | 40,93.02 449 | 30,89.08 450 | 50,95.72 451 | 10,34.54 452 | 20,73.15 453 | 30,90.07 454 | 40,91.44 455 | 50,95.01 456 | 10,32.34 457 | 20,86.94 458 | 30,89.18 459 | 40,90.96 460 | 50,94.08 461 | 10,69.43 462 | 20,76.51 463 | 30,88.78 464 | 40,94.20 465 | 50,90.34 466 | 20,84.16 467 | 10,64.95 468 | 30,84.42 469 | 40,89.88 470 | 50,95.58 471 | 10,54.50 472 | 20,86.11 473 | 30,92.43 474 | 40,92.69 475 | 50,93.43 476 | 10,67.23 477 | 20,84.14 478 | 30,93.65 479 | 40,92.69 480 | 50,94.43 481 | 10,34.78 482 | 20,89.52 483 | 30,93.92 484 | 40,93.51 485 | 50,95.05 486 | 10,53.01 487 | 20,72.42 488 | 30,86.82 489 | 40,92.12 490 | 50,93.82 491 | 10,33.73 492 | 20,90.22 493 | 30,92.57 494 | 40,92.07 495 | 50,94.67 496 | 20,84.42 497 | 10,67.31 498 | 30,92.97 499 | 40,93.47 500 | 50,95.02 501 | -------------------------------------------------------------------------------- /ResultsFromPaper/MNIST_vae_k12_poisson_accuracy50trials.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-05_06:44 2 | 10,92.81 3 | 20,94.69 4 | 50,95.86 5 | 20,94.23 6 | 10,91.33 7 | 40,95.66 8 | 30,94.77 9 | 40,95.38 10 | 20,92.42 11 | 10,90.89 12 | 50,94.99 13 | 30,95.01 14 | 40,95.11 15 | 30,93.03 16 | 50,96.03 17 | 20,95.51 18 | 10,94.74 19 | 50,95.94 20 | 40,95.25 21 | 20,91.79 22 | 30,95.08 23 | 10,93.00 24 | 40,93.67 25 | 10,91.96 26 | 30,93.98 27 | 50,95.65 28 | 20,93.72 29 | 40,94.74 30 | 30,94.92 31 | 30,95.41 32 | 40,95.24 33 | 50,96.23 34 | 20,95.75 35 | 10,81.92 36 | 30,94.65 37 | 50,95.23 38 | 10,91.63 39 | 50,95.72 40 | 50,95.43 41 | 10,92.43 42 | 40,94.71 43 | 20,91.53 44 | 10,85.24 45 | 20,94.81 46 | 20,94.84 47 | 40,95.17 48 | 30,95.17 49 | 30,95.36 50 | 50,94.44 51 | 40,95.64 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/MNIST_vae_k14_poisson_accuracy50trials.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-05_06:48 2 | 10,91.05 3 | 10,92.80 4 | 50,95.83 5 | 20,94.10 6 | 30,94.73 7 | 40,95.56 8 | 20,92.25 9 | 20,94.64 10 | 40,95.39 11 | 10,90.70 12 | 50,94.92 13 | 30,94.99 14 | 50,95.98 15 | 20,91.75 16 | 40,95.14 17 | 30,93.97 18 | 30,92.86 19 | 10,94.59 20 | 50,96.03 21 | 10,92.84 22 | 20,95.46 23 | 30,94.99 24 | 40,95.24 25 | 40,94.03 26 | 30,94.78 27 | 50,95.58 28 | 40,95.14 29 | 20,93.46 30 | 50,96.18 31 | 10,91.77 32 | 50,95.77 33 | 20,95.68 34 | 10,81.67 35 | 40,94.69 36 | 30,95.30 37 | 10,91.51 38 | 20,94.80 39 | 10,92.32 40 | 30,94.69 41 | 50,95.18 42 | 40,94.86 43 | 20,94.76 44 | 30,95.03 45 | 20,91.68 46 | 50,95.39 47 | 10,85.04 48 | 40,95.17 49 | 50,94.43 50 | 30,95.34 51 | 40,95.68 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/MNIST_vae_k16_poisson_accuracy50trials.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-08_13:37 2 | 20,94.03 3 | 10,92.75 4 | 50,95.80 5 | 30,94.64 6 | 10,90.89 7 | 40,95.47 8 | 20,91.99 9 | 20,94.55 10 | 40,95.33 11 | 10,90.67 12 | 50,94.92 13 | 30,94.80 14 | 50,95.94 15 | 50,96.05 16 | 20,91.65 17 | 40,95.03 18 | 30,94.90 19 | 30,92.76 20 | 10,94.35 21 | 10,92.69 22 | 20,95.43 23 | 40,95.20 24 | 40,94.21 25 | 20,93.12 26 | 30,93.83 27 | 50,95.56 28 | 10,91.71 29 | 40,94.57 30 | 30,94.76 31 | 30,95.25 32 | 50,96.14 33 | 40,95.09 34 | 20,95.70 35 | 10,81.55 36 | 50,95.14 37 | 30,94.66 38 | 50,95.68 39 | 10,91.33 40 | 10,92.30 41 | 20,91.71 42 | 50,95.36 43 | 40,94.78 44 | 10,85.09 45 | 20,94.70 46 | 30,94.88 47 | 20,94.73 48 | 40,95.16 49 | 50,94.30 50 | 30,95.30 51 | 40,95.72 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/MNIST_vae_k18_poisson_accuracy50trials.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-05_06:52 2 | 10,92.68 3 | 30,94.66 4 | 50,95.74 5 | 30,94.53 6 | 20,93.78 7 | 40,95.39 8 | 10,90.60 9 | 20,94.45 10 | 10,90.36 11 | 40,95.29 12 | 50,94.80 13 | 20,91.75 14 | 40,95.12 15 | 20,95.37 16 | 40,95.01 17 | 50,95.87 18 | 30,92.81 19 | 10,94.06 20 | 50,96.01 21 | 10,92.48 22 | 20,91.46 23 | 30,93.80 24 | 40,94.26 25 | 30,94.82 26 | 50,95.46 27 | 20,95.59 28 | 30,95.12 29 | 10,91.35 30 | 30,94.71 31 | 20,93.07 32 | 50,96.07 33 | 40,95.04 34 | 10,80.91 35 | 40,94.58 36 | 50,95.63 37 | 10,91.11 38 | 50,95.31 39 | 50,95.03 40 | 30,94.58 41 | 10,92.25 42 | 30,95.25 43 | 20,91.66 44 | 40,94.69 45 | 20,94.68 46 | 40,95.09 47 | 10,84.98 48 | 50,94.28 49 | 20,94.65 50 | 40,95.70 51 | 30,94.72 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/MNIST_vae_k20_poisson_accuracy50trials.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-05_06:56 2 | 10,92.63 3 | 50,95.66 4 | 30,94.45 5 | 10,90.04 6 | 40,95.30 7 | 20,93.47 8 | 20,94.28 9 | 40,95.19 10 | 50,94.76 11 | 20,91.61 12 | 30,94.57 13 | 30,94.76 14 | 40,94.92 15 | 50,95.77 16 | 10,90.21 17 | 20,95.29 18 | 10,93.83 19 | 50,95.96 20 | 40,94.16 21 | 30,92.79 22 | 10,92.48 23 | 20,91.35 24 | 40,95.11 25 | 10,91.03 26 | 30,93.71 27 | 50,95.40 28 | 40,94.95 29 | 30,94.57 30 | 20,92.97 31 | 50,96.02 32 | 40,94.50 33 | 20,95.56 34 | 10,80.51 35 | 20,94.62 36 | 10,90.92 37 | 30,95.02 38 | 10,92.09 39 | 50,95.58 40 | 50,95.03 41 | 30,94.53 42 | 40,95.00 43 | 40,94.60 44 | 20,94.56 45 | 20,91.51 46 | 50,95.21 47 | 30,94.63 48 | 50,94.27 49 | 10,84.99 50 | 30,95.17 51 | 40,95.72 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/MNIST_vae_k6_poisson_accuracy50trials.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-05_06:33 2 | 10,92.22 3 | 20,94.78 4 | 50,95.91 5 | 20,94.48 6 | 40,95.17 7 | 20,92.80 8 | 10,92.71 9 | 30,94.78 10 | 40,95.76 11 | 10,91.81 12 | 50,95.09 13 | 30,95.42 14 | 10,95.15 15 | 40,94.79 16 | 50,95.92 17 | 30,94.98 18 | 30,93.94 19 | 20,92.49 20 | 50,95.46 21 | 30,92.68 22 | 40,95.26 23 | 10,92.42 24 | 50,95.68 25 | 10,92.99 26 | 20,95.63 27 | 20,94.72 28 | 40,93.45 29 | 30,95.09 30 | 30,95.43 31 | 40,95.32 32 | 50,96.15 33 | 20,95.78 34 | 10,82.40 35 | 10,92.03 36 | 50,95.58 37 | 40,94.87 38 | 30,94.88 39 | 20,95.07 40 | 40,94.35 41 | 50,95.50 42 | 30,95.21 43 | 50,95.46 44 | 10,92.32 45 | 20,90.98 46 | 20,94.77 47 | 10,85.63 48 | 50,94.69 49 | 40,94.86 50 | 40,95.19 51 | 30,95.44 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/MNIST_vae_k8_poisson_accuracy50trials.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-06-05_06:39 2 | 10,91.65 3 | 30,94.84 4 | 50,95.93 5 | 20,94.54 6 | 20,94.70 7 | 10,92.86 8 | 40,95.39 9 | 40,95.79 10 | 20,92.66 11 | 30,95.26 12 | 50,94.95 13 | 10,91.43 14 | 40,95.03 15 | 20,95.65 16 | 10,94.93 17 | 10,93.15 18 | 50,96.07 19 | 50,95.78 20 | 30,92.95 21 | 40,95.26 22 | 20,92.16 23 | 30,95.03 24 | 10,92.40 25 | 30,94.21 26 | 40,93.39 27 | 50,95.74 28 | 20,94.38 29 | 30,95.12 30 | 40,95.37 31 | 30,95.49 32 | 20,95.76 33 | 50,96.25 34 | 50,95.69 35 | 10,81.68 36 | 40,94.90 37 | 50,95.41 38 | 10,92.09 39 | 30,94.87 40 | 10,92.51 41 | 10,85.52 42 | 40,94.64 43 | 20,91.40 44 | 30,95.50 45 | 50,95.45 46 | 20,95.04 47 | 40,95.05 48 | 20,95.09 49 | 50,94.25 50 | 30,95.32 51 | 40,95.45 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/cifarModRate_aet_k10_p10laplace_none_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-07-17_12:47 2 | 1600,69.21 3 | 1600,68.61 4 | 800,66.29 5 | 800,66.95 6 | 400,63.05 7 | 400,63.25 8 | 200,59.80 9 | 200,60.22 10 | 100,56.46 11 | 1600,68.61 12 | 800,66.40 13 | 400,62.79 14 | 100,57.03 15 | 200,58.37 16 | 400,63.72 17 | 1600,68.86 18 | 100,57.67 19 | 800,65.72 20 | 200,61.54 21 | 400,64.21 22 | 800,66.28 23 | 100,56.84 24 | 1600,68.74 25 | 200,62.11 26 | 100,52.08 27 | 800,66.99 28 | 1600,68.90 29 | 400,64.54 30 | 200,61.92 31 | 800,65.35 32 | 400,64.90 33 | 1600,68.67 34 | 100,57.83 35 | 200,61.48 36 | 800,67.17 37 | 100,57.16 38 | 400,63.55 39 | 1600,68.67 40 | 200,59.77 41 | 100,54.20 42 | 1600,68.21 43 | 800,66.59 44 | 400,63.59 45 | 200,58.95 46 | 800,65.60 47 | 100,57.30 48 | 400,64.21 49 | 1600,68.45 50 | 200,60.02 51 | 100,57.70 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/cifarModRate_aet_k10_sparselabelpropagation_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-07-16_23:22 2 | 200,18.53 3 | 100,20.55 4 | 200,18.60 5 | 1600,13.60 6 | 100,14.16 7 | 1600,13.62 8 | 400,17.99 9 | 800,23.35 10 | 800,20.69 11 | 400,19.17 12 | 400,17.32 13 | 800,22.61 14 | 200,12.33 15 | 100,10.81 16 | 1600,16.07 17 | 100,14.83 18 | 400,18.51 19 | 200,25.02 20 | 800,20.91 21 | 1600,13.68 22 | 100,17.30 23 | 800,20.59 24 | 400,23.40 25 | 200,18.89 26 | 200,18.02 27 | 100,15.94 28 | 1600,14.30 29 | 400,21.86 30 | 100,10.39 31 | 800,19.76 32 | 1600,14.92 33 | 200,21.36 34 | 400,20.28 35 | 800,22.75 36 | 1600,15.55 37 | 400,21.71 38 | 100,15.85 39 | 200,12.38 40 | 100,19.02 41 | 400,18.63 42 | 800,23.67 43 | 1600,16.06 44 | 200,12.92 45 | 1600,16.02 46 | 800,20.95 47 | 100,17.36 48 | 400,21.36 49 | 800,22.00 50 | 200,15.95 51 | 1600,16.39 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/cifarMult1_aet_k10_modularitymbo_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2021-03-22_23:02 2 | 10,46.09 3 | 10,36.77 4 | 10,48.52 5 | 10,45.11 6 | 10,45.51 7 | 10,37.82 8 | 10,32.84 9 | 10,50.50 10 | 10,34.50 11 | 10,46.05 12 | 10,48.61 13 | 10,41.92 14 | 10,45.97 15 | 10,39.70 16 | 10,29.78 17 | 10,40.52 18 | 10,41.94 19 | 10,39.72 20 | 10,41.33 21 | 10,26.40 22 | 10,50.92 23 | 10,39.77 24 | 10,21.24 25 | 10,46.32 26 | 10,29.61 27 | 10,43.10 28 | 10,43.65 29 | 10,43.91 30 | 10,49.24 31 | 10,43.26 32 | 10,37.87 33 | 10,33.09 34 | 10,37.15 35 | 10,26.05 36 | 10,34.16 37 | 10,20.41 38 | 10,32.74 39 | 10,26.62 40 | 10,45.91 41 | 10,33.00 42 | 10,28.32 43 | 10,35.99 44 | 10,47.02 45 | 10,39.41 46 | 10,35.87 47 | 10,47.26 48 | 10,43.27 49 | 10,34.72 50 | 10,45.79 51 | 10,35.50 52 | 10,46.38 53 | 10,46.99 54 | 10,39.61 55 | 10,44.51 56 | 10,47.39 57 | 10,40.96 58 | 10,43.63 59 | 10,50.71 60 | 10,41.30 61 | 10,35.50 62 | 10,40.80 63 | 10,43.35 64 | 10,40.38 65 | 10,41.88 66 | 10,52.59 67 | 10,45.33 68 | 10,41.48 69 | 10,38.71 70 | 10,38.45 71 | 10,44.81 72 | 10,48.40 73 | 10,28.58 74 | 10,43.68 75 | 10,36.90 76 | 10,32.64 77 | 10,16.16 78 | 10,47.85 79 | 10,36.68 80 | 10,33.49 81 | 10,39.68 82 | 10,52.27 83 | 10,34.01 84 | 10,47.24 85 | 10,49.49 86 | 10,38.39 87 | 10,42.41 88 | 10,33.31 89 | 10,35.91 90 | 10,44.99 91 | 10,31.29 92 | 10,44.00 93 | 10,22.07 94 | 10,34.65 95 | 10,41.39 96 | 10,32.81 97 | 10,37.00 98 | 10,29.89 99 | 10,32.99 100 | 10,47.08 101 | 10,41.44 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/cifarMult1_aet_k10_poisson_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-07-16_18:33 2 | 10,42.21 3 | 10,37.33 4 | 10,45.27 5 | 10,50.95 6 | 10,39.99 7 | 10,38.60 8 | 10,38.48 9 | 10,44.46 10 | 10,47.68 11 | 10,43.74 12 | 10,49.99 13 | 10,35.27 14 | 10,42.65 15 | 10,39.07 16 | 10,43.64 17 | 10,40.25 18 | 10,34.12 19 | 10,39.57 20 | 10,39.20 21 | 10,35.39 22 | 10,48.47 23 | 10,39.83 24 | 10,39.93 25 | 10,20.71 26 | 10,47.88 27 | 10,44.08 28 | 10,45.31 29 | 10,29.73 30 | 10,41.13 31 | 10,44.16 32 | 10,31.57 33 | 10,41.50 34 | 10,38.03 35 | 10,40.70 36 | 10,32.83 37 | 10,42.56 38 | 10,40.09 39 | 10,32.15 40 | 10,21.99 41 | 10,35.95 42 | 10,36.09 43 | 10,47.01 44 | 10,43.65 45 | 10,38.73 46 | 10,45.99 47 | 10,43.21 48 | 10,46.28 49 | 10,44.47 50 | 10,48.72 51 | 10,41.06 52 | 10,52.14 53 | 10,38.85 54 | 10,48.60 55 | 10,49.00 56 | 10,36.70 57 | 10,49.35 58 | 10,47.40 59 | 10,48.20 60 | 10,43.30 61 | 10,50.69 62 | 10,43.79 63 | 10,44.61 64 | 10,39.38 65 | 10,42.12 66 | 10,48.67 67 | 10,42.90 68 | 10,45.68 69 | 10,43.33 70 | 10,41.04 71 | 10,38.33 72 | 10,46.82 73 | 10,45.86 74 | 10,33.21 75 | 10,40.83 76 | 10,37.42 77 | 10,44.83 78 | 10,24.63 79 | 10,42.02 80 | 10,42.08 81 | 10,46.32 82 | 10,37.79 83 | 10,38.33 84 | 10,49.19 85 | 10,34.43 86 | 10,47.14 87 | 10,37.66 88 | 10,39.94 89 | 10,38.36 90 | 10,28.20 91 | 10,34.21 92 | 10,45.40 93 | 10,35.14 94 | 10,47.83 95 | 10,42.65 96 | 10,28.82 97 | 10,40.46 98 | 10,38.89 99 | 10,43.24 100 | 10,41.52 101 | 10,45.38 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/cifarMult2_aet_k10_modularitymbo_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2021-03-22_23:03 2 | 15,45.98 3 | 15,40.33 4 | 15,43.59 5 | 15,48.15 6 | 15,48.14 7 | 15,40.87 8 | 15,46.44 9 | 15,39.15 10 | 15,39.73 11 | 15,33.61 12 | 15,34.80 13 | 15,44.61 14 | 15,49.76 15 | 15,45.94 16 | 15,43.70 17 | 15,35.48 18 | 15,44.45 19 | 15,32.26 20 | 15,36.29 21 | 15,51.71 22 | 15,38.69 23 | 15,45.49 24 | 15,45.03 25 | 15,41.15 26 | 15,33.28 27 | 15,45.00 28 | 15,37.77 29 | 15,36.00 30 | 15,43.49 31 | 15,39.57 32 | 15,47.81 33 | 15,27.00 34 | 15,27.15 35 | 15,26.11 36 | 15,32.74 37 | 15,44.55 38 | 15,45.51 39 | 15,33.27 40 | 15,42.83 41 | 15,44.19 42 | 15,44.23 43 | 15,44.87 44 | 15,32.65 45 | 15,40.24 46 | 15,38.95 47 | 15,36.75 48 | 15,38.27 49 | 15,40.97 50 | 15,40.20 51 | 15,41.54 52 | 15,26.51 53 | 15,48.08 54 | 15,42.26 55 | 15,44.74 56 | 15,44.05 57 | 15,36.45 58 | 15,35.57 59 | 15,42.69 60 | 15,47.99 61 | 15,28.85 62 | 15,42.65 63 | 15,46.28 64 | 15,42.75 65 | 15,30.40 66 | 15,39.92 67 | 15,37.46 68 | 15,48.76 69 | 15,40.20 70 | 15,40.63 71 | 15,35.96 72 | 15,48.18 73 | 15,38.00 74 | 15,44.64 75 | 15,44.54 76 | 15,40.32 77 | 15,40.22 78 | 15,43.28 79 | 15,48.59 80 | 15,44.15 81 | 15,49.72 82 | 15,51.55 83 | 15,36.67 84 | 15,34.16 85 | 15,40.62 86 | 15,44.93 87 | 15,37.32 88 | 15,48.33 89 | 15,47.50 90 | 15,48.17 91 | 15,40.85 92 | 15,42.88 93 | 15,42.89 94 | 15,36.52 95 | 15,42.90 96 | 15,32.31 97 | 15,31.30 98 | 15,37.31 99 | 15,36.99 100 | 15,36.25 101 | 15,45.70 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/cifarMult2_aet_k10_poisson_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-07-16_18:37 2 | 15,43.74 3 | 15,50.44 4 | 15,41.85 5 | 15,45.00 6 | 15,33.27 7 | 15,48.94 8 | 15,50.50 9 | 15,42.73 10 | 15,44.18 11 | 15,41.29 12 | 15,44.45 13 | 15,34.06 14 | 15,44.40 15 | 15,43.60 16 | 15,38.41 17 | 15,48.12 18 | 15,35.63 19 | 15,41.02 20 | 15,41.07 21 | 15,45.69 22 | 15,50.12 23 | 15,46.37 24 | 15,46.66 25 | 15,49.74 26 | 15,43.64 27 | 15,42.33 28 | 15,44.37 29 | 15,43.69 30 | 15,44.78 31 | 15,41.07 32 | 15,49.58 33 | 15,34.86 34 | 15,29.41 35 | 15,32.52 36 | 15,49.66 37 | 15,29.39 38 | 15,45.36 39 | 15,40.08 40 | 15,42.55 41 | 15,42.37 42 | 15,50.33 43 | 15,35.14 44 | 15,41.95 45 | 15,35.17 46 | 15,40.66 47 | 15,44.04 48 | 15,44.87 49 | 15,46.15 50 | 15,41.81 51 | 15,44.19 52 | 15,52.10 53 | 15,37.08 54 | 15,41.00 55 | 15,43.71 56 | 15,41.29 57 | 15,42.73 58 | 15,37.80 59 | 15,37.67 60 | 15,43.77 61 | 15,34.05 62 | 15,38.69 63 | 15,48.25 64 | 15,47.10 65 | 15,41.85 66 | 15,43.20 67 | 15,44.97 68 | 15,43.89 69 | 15,41.47 70 | 15,37.31 71 | 15,46.84 72 | 15,36.39 73 | 15,42.64 74 | 15,49.41 75 | 15,48.90 76 | 15,43.06 77 | 15,42.54 78 | 15,48.53 79 | 15,42.84 80 | 15,48.00 81 | 15,48.95 82 | 15,47.71 83 | 15,39.97 84 | 15,42.34 85 | 15,50.26 86 | 15,42.74 87 | 15,51.56 88 | 15,37.00 89 | 15,54.41 90 | 15,45.12 91 | 15,42.03 92 | 15,43.41 93 | 15,44.10 94 | 15,45.04 95 | 15,42.05 96 | 15,32.72 97 | 15,34.16 98 | 15,39.76 99 | 15,45.38 100 | 15,40.28 101 | 15,44.44 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/cifarMult3_aet_k10_modularitymbo_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2021-03-22_23:03 2 | 20,45.46 3 | 20,34.51 4 | 20,46.45 5 | 20,38.46 6 | 20,50.40 7 | 20,40.51 8 | 20,51.25 9 | 20,39.43 10 | 20,40.76 11 | 20,39.24 12 | 20,24.31 13 | 20,44.78 14 | 20,37.45 15 | 20,39.97 16 | 20,48.99 17 | 20,41.58 18 | 20,37.68 19 | 20,42.00 20 | 20,37.73 21 | 20,50.04 22 | 20,42.47 23 | 20,38.64 24 | 20,54.64 25 | 20,51.93 26 | 20,37.03 27 | 20,43.36 28 | 20,47.49 29 | 20,41.22 30 | 20,52.15 31 | 20,51.16 32 | 20,38.35 33 | 20,44.40 34 | 20,42.08 35 | 20,41.20 36 | 20,43.23 37 | 20,49.00 38 | 20,39.95 39 | 20,39.06 40 | 20,54.49 41 | 20,50.72 42 | 20,48.03 43 | 20,37.93 44 | 20,44.22 45 | 20,40.64 46 | 20,46.89 47 | 20,39.26 48 | 20,39.08 49 | 20,50.92 50 | 20,43.38 51 | 20,42.12 52 | 20,38.32 53 | 20,38.52 54 | 20,40.16 55 | 20,34.72 56 | 20,51.11 57 | 20,45.79 58 | 20,42.09 59 | 20,42.05 60 | 20,46.05 61 | 20,38.59 62 | 20,45.05 63 | 20,40.11 64 | 20,38.00 65 | 20,45.84 66 | 20,34.88 67 | 20,39.96 68 | 20,46.02 69 | 20,52.12 70 | 20,44.49 71 | 20,35.13 72 | 20,40.28 73 | 20,52.11 74 | 20,40.43 75 | 20,51.47 76 | 20,38.44 77 | 20,35.61 78 | 20,37.33 79 | 20,41.47 80 | 20,45.40 81 | 20,48.67 82 | 20,43.26 83 | 20,33.48 84 | 20,33.54 85 | 20,42.58 86 | 20,41.91 87 | 20,49.09 88 | 20,47.65 89 | 20,46.83 90 | 20,45.80 91 | 20,22.64 92 | 20,49.00 93 | 20,52.25 94 | 20,41.89 95 | 20,47.38 96 | 20,46.87 97 | 20,52.96 98 | 20,40.15 99 | 20,55.32 100 | 20,40.42 101 | 20,47.91 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/cifarMult3_aet_k10_poisson_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-07-16_18:42 2 | 20,48.12 3 | 20,43.88 4 | 20,45.55 5 | 20,50.85 6 | 20,41.78 7 | 20,41.48 8 | 20,44.21 9 | 20,47.91 10 | 20,41.73 11 | 20,45.90 12 | 20,37.97 13 | 20,34.70 14 | 20,42.64 15 | 20,47.15 16 | 20,53.06 17 | 20,46.42 18 | 20,41.87 19 | 20,46.19 20 | 20,37.41 21 | 20,48.17 22 | 20,55.64 23 | 20,43.29 24 | 20,47.44 25 | 20,48.85 26 | 20,50.77 27 | 20,44.63 28 | 20,53.07 29 | 20,48.09 30 | 20,54.23 31 | 20,48.42 32 | 20,43.06 33 | 20,45.71 34 | 20,44.12 35 | 20,46.03 36 | 20,48.60 37 | 20,42.28 38 | 20,46.69 39 | 20,43.14 40 | 20,54.80 41 | 20,50.33 42 | 20,48.95 43 | 20,49.15 44 | 20,48.80 45 | 20,38.24 46 | 20,46.43 47 | 20,45.58 48 | 20,42.46 49 | 20,40.45 50 | 20,44.04 51 | 20,46.69 52 | 20,43.76 53 | 20,42.59 54 | 20,39.33 55 | 20,33.61 56 | 20,51.32 57 | 20,48.05 58 | 20,47.28 59 | 20,51.64 60 | 20,42.07 61 | 20,39.91 62 | 20,43.58 63 | 20,49.51 64 | 20,38.26 65 | 20,45.05 66 | 20,46.03 67 | 20,45.39 68 | 20,51.47 69 | 20,42.63 70 | 20,47.24 71 | 20,36.32 72 | 20,32.88 73 | 20,49.63 74 | 20,51.56 75 | 20,43.26 76 | 20,47.61 77 | 20,43.03 78 | 20,44.75 79 | 20,44.93 80 | 20,44.78 81 | 20,49.29 82 | 20,49.28 83 | 20,44.33 84 | 20,35.08 85 | 20,43.40 86 | 20,54.02 87 | 20,46.56 88 | 20,50.00 89 | 20,46.73 90 | 20,40.54 91 | 20,50.61 92 | 20,37.04 93 | 20,49.24 94 | 20,49.52 95 | 20,47.55 96 | 20,48.44 97 | 20,43.74 98 | 20,49.34 99 | 20,46.30 100 | 20,48.87 101 | 20,54.87 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/cifarMult4_aet_k10_modularitymbo_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2021-03-22_23:04 2 | 25,41.46 3 | 25,42.87 4 | 25,49.86 5 | 25,46.45 6 | 25,49.06 7 | 25,42.99 8 | 25,42.08 9 | 25,38.59 10 | 25,47.72 11 | 25,33.44 12 | 25,40.58 13 | 25,47.33 14 | 25,41.22 15 | 25,35.71 16 | 25,42.08 17 | 25,45.51 18 | 25,41.19 19 | 25,54.65 20 | 25,45.06 21 | 25,48.68 22 | 25,42.75 23 | 25,49.89 24 | 25,47.49 25 | 25,50.22 26 | 25,43.68 27 | 25,45.19 28 | 25,43.32 29 | 25,42.79 30 | 25,38.24 31 | 25,37.50 32 | 25,52.02 33 | 25,38.53 34 | 25,47.56 35 | 25,35.64 36 | 25,41.48 37 | 25,41.44 38 | 25,45.52 39 | 25,43.95 40 | 25,45.16 41 | 25,45.88 42 | 25,51.67 43 | 25,41.64 44 | 25,42.16 45 | 25,50.06 46 | 25,45.41 47 | 25,46.36 48 | 25,43.72 49 | 25,48.52 50 | 25,44.10 51 | 25,41.92 52 | 25,46.43 53 | 25,45.71 54 | 25,47.18 55 | 25,42.81 56 | 25,52.73 57 | 25,51.10 58 | 25,47.24 59 | 25,45.60 60 | 25,47.73 61 | 25,41.89 62 | 25,44.36 63 | 25,38.69 64 | 25,37.29 65 | 25,42.23 66 | 25,42.52 67 | 25,47.08 68 | 25,50.38 69 | 25,46.45 70 | 25,47.73 71 | 25,41.32 72 | 25,53.69 73 | 25,46.31 74 | 25,36.74 75 | 25,43.32 76 | 25,29.88 77 | 25,39.53 78 | 25,47.70 79 | 25,41.54 80 | 25,41.79 81 | 25,40.85 82 | 25,41.16 83 | 25,37.56 84 | 25,48.73 85 | 25,45.85 86 | 25,33.93 87 | 25,33.53 88 | 25,33.77 89 | 25,34.54 90 | 25,44.12 91 | 25,45.70 92 | 25,51.79 93 | 25,46.37 94 | 25,46.91 95 | 25,42.87 96 | 25,47.46 97 | 25,48.89 98 | 25,51.86 99 | 25,39.14 100 | 25,40.18 101 | 25,33.04 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/cifarMult4_aet_k10_poisson_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-07-16_18:47 2 | 25,47.22 3 | 25,43.93 4 | 25,46.62 5 | 25,45.98 6 | 25,50.30 7 | 25,47.91 8 | 25,38.27 9 | 25,43.17 10 | 25,54.29 11 | 25,39.64 12 | 25,39.02 13 | 25,47.70 14 | 25,53.22 15 | 25,36.59 16 | 25,45.94 17 | 25,44.87 18 | 25,47.63 19 | 25,49.81 20 | 25,40.52 21 | 25,50.07 22 | 25,44.35 23 | 25,48.87 24 | 25,50.55 25 | 25,44.06 26 | 25,50.15 27 | 25,52.95 28 | 25,41.22 29 | 25,45.16 30 | 25,42.29 31 | 25,42.21 32 | 25,51.99 33 | 25,43.62 34 | 25,45.80 35 | 25,47.21 36 | 25,47.74 37 | 25,46.77 38 | 25,46.88 39 | 25,48.74 40 | 25,47.70 41 | 25,47.09 42 | 25,43.49 43 | 25,45.02 44 | 25,45.92 45 | 25,47.51 46 | 25,50.01 47 | 25,51.33 48 | 25,44.70 49 | 25,47.05 50 | 25,46.05 51 | 25,48.39 52 | 25,54.51 53 | 25,46.67 54 | 25,47.65 55 | 25,50.97 56 | 25,51.84 57 | 25,48.54 58 | 25,48.54 59 | 25,51.90 60 | 25,51.46 61 | 25,43.53 62 | 25,41.72 63 | 25,42.80 64 | 25,37.98 65 | 25,41.52 66 | 25,40.74 67 | 25,46.84 68 | 25,47.17 69 | 25,40.25 70 | 25,48.46 71 | 25,48.90 72 | 25,49.68 73 | 25,42.09 74 | 25,44.65 75 | 25,43.34 76 | 25,39.77 77 | 25,43.32 78 | 25,49.31 79 | 25,47.25 80 | 25,41.24 81 | 25,50.35 82 | 25,45.35 83 | 25,47.06 84 | 25,46.09 85 | 25,47.80 86 | 25,36.29 87 | 25,41.46 88 | 25,36.64 89 | 25,39.86 90 | 25,52.07 91 | 25,45.69 92 | 25,46.61 93 | 25,48.14 94 | 25,47.43 95 | 25,52.53 96 | 25,47.56 97 | 25,47.26 98 | 25,45.23 99 | 25,42.86 100 | 25,41.01 101 | 25,44.09 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/cifarMult5_aet_k10_modularitymbo_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2021-03-22_23:04 2 | 30,51.76 3 | 30,47.60 4 | 30,34.24 5 | 30,37.37 6 | 30,42.44 7 | 30,51.15 8 | 30,44.88 9 | 30,33.96 10 | 30,49.88 11 | 30,44.95 12 | 30,45.77 13 | 30,42.19 14 | 30,50.20 15 | 30,49.32 16 | 30,50.52 17 | 30,47.90 18 | 30,48.12 19 | 30,38.09 20 | 30,50.04 21 | 30,48.26 22 | 30,44.93 23 | 30,45.65 24 | 30,39.91 25 | 30,44.21 26 | 30,41.72 27 | 30,41.85 28 | 30,48.07 29 | 30,36.28 30 | 30,46.37 31 | 30,44.70 32 | 30,38.21 33 | 30,38.88 34 | 30,48.03 35 | 30,36.99 36 | 30,47.88 37 | 30,45.72 38 | 30,43.03 39 | 30,47.54 40 | 30,50.26 41 | 30,38.18 42 | 30,48.26 43 | 30,43.46 44 | 30,46.37 45 | 30,52.16 46 | 30,46.53 47 | 30,38.58 48 | 30,40.48 49 | 30,50.41 50 | 30,45.70 51 | 30,36.09 52 | 30,49.09 53 | 30,43.36 54 | 30,41.59 55 | 30,53.08 56 | 30,41.64 57 | 30,53.98 58 | 30,49.53 59 | 30,44.24 60 | 30,42.38 61 | 30,47.99 62 | 30,50.82 63 | 30,35.92 64 | 30,42.38 65 | 30,37.82 66 | 30,50.40 67 | 30,46.62 68 | 30,51.88 69 | 30,47.36 70 | 30,49.74 71 | 30,55.85 72 | 30,48.86 73 | 30,35.79 74 | 30,43.41 75 | 30,30.25 76 | 30,50.26 77 | 30,35.51 78 | 30,47.24 79 | 30,41.50 80 | 30,39.78 81 | 30,51.39 82 | 30,43.55 83 | 30,46.66 84 | 30,38.48 85 | 30,48.14 86 | 30,42.38 87 | 30,49.65 88 | 30,42.77 89 | 30,49.55 90 | 30,41.22 91 | 30,36.61 92 | 30,49.31 93 | 30,37.72 94 | 30,39.78 95 | 30,45.35 96 | 30,39.02 97 | 30,50.81 98 | 30,46.65 99 | 30,39.41 100 | 30,34.91 101 | 30,51.85 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/cifarMult5_aet_k10_poisson_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-07-16_18:51 2 | 30,47.81 3 | 30,45.08 4 | 30,53.23 5 | 30,50.13 6 | 30,51.41 7 | 30,50.42 8 | 30,54.45 9 | 30,50.10 10 | 30,43.55 11 | 30,53.07 12 | 30,51.23 13 | 30,52.37 14 | 30,52.51 15 | 30,46.91 16 | 30,46.85 17 | 30,44.58 18 | 30,52.76 19 | 30,45.56 20 | 30,43.59 21 | 30,47.74 22 | 30,45.31 23 | 30,45.60 24 | 30,46.97 25 | 30,43.06 26 | 30,45.19 27 | 30,46.70 28 | 30,46.75 29 | 30,49.58 30 | 30,51.95 31 | 30,41.79 32 | 30,51.77 33 | 30,47.72 34 | 30,46.00 35 | 30,40.60 36 | 30,50.31 37 | 30,44.74 38 | 30,49.69 39 | 30,49.72 40 | 30,39.67 41 | 30,44.23 42 | 30,32.47 43 | 30,52.24 44 | 30,51.14 45 | 30,44.94 46 | 30,47.77 47 | 30,45.47 48 | 30,43.69 49 | 30,50.81 50 | 30,46.66 51 | 30,43.36 52 | 30,51.87 53 | 30,47.47 54 | 30,46.95 55 | 30,46.99 56 | 30,43.34 57 | 30,52.40 58 | 30,42.46 59 | 30,54.64 60 | 30,55.30 61 | 30,41.79 62 | 30,52.01 63 | 30,42.40 64 | 30,52.22 65 | 30,52.16 66 | 30,49.14 67 | 30,54.71 68 | 30,47.58 69 | 30,44.12 70 | 30,49.97 71 | 30,54.95 72 | 30,55.40 73 | 30,49.55 74 | 30,45.10 75 | 30,39.18 76 | 30,39.37 77 | 30,37.74 78 | 30,44.69 79 | 30,51.44 80 | 30,46.44 81 | 30,43.73 82 | 30,53.31 83 | 30,45.91 84 | 30,48.58 85 | 30,49.31 86 | 30,46.30 87 | 30,36.60 88 | 30,53.66 89 | 30,38.79 90 | 30,46.86 91 | 30,51.55 92 | 30,50.17 93 | 30,47.70 94 | 30,44.05 95 | 30,47.11 96 | 30,51.90 97 | 30,45.91 98 | 30,45.32 99 | 30,42.92 100 | 30,44.83 101 | 30,39.96 102 | -------------------------------------------------------------------------------- /ResultsFromPaper/cifar_aet_k10_p10laplace_none_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-07-17_10:58 2 | 50,44.58 3 | 50,48.02 4 | 40,46.95 5 | 40,46.26 6 | 30,40.52 7 | 30,45.61 8 | 20,35.09 9 | 20,47.30 10 | 10,28.59 11 | 10,30.56 12 | 40,47.09 13 | 50,55.83 14 | 30,44.20 15 | 20,36.60 16 | 30,42.18 17 | 20,38.43 18 | 40,49.38 19 | 10,22.83 20 | 50,53.52 21 | 10,29.50 22 | 30,47.62 23 | 40,46.98 24 | 50,51.04 25 | 20,35.53 26 | 10,26.48 27 | 40,46.27 28 | 50,49.13 29 | 20,28.82 30 | 10,12.78 31 | 30,42.29 32 | 40,49.63 33 | 50,49.41 34 | 20,37.47 35 | 30,41.39 36 | 50,47.14 37 | 40,50.55 38 | 30,39.04 39 | 20,30.94 40 | 10,28.28 41 | 10,15.66 42 | 30,42.56 43 | 40,53.85 44 | 20,27.30 45 | 50,54.22 46 | 10,30.02 47 | 40,44.17 48 | 30,35.91 49 | 50,43.75 50 | 20,32.54 51 | 10,35.46 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/cifar_aet_k10_poisson_accuracy50trials.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-07-20_14:57 2 | 10,40.90 3 | 20,48.17 4 | 30,52.57 5 | 10,52.34 6 | 10,30.80 7 | 50,51.77 8 | 40,53.71 9 | 30,53.72 10 | 20,47.70 11 | 50,52.19 12 | 40,52.10 13 | 20,48.95 14 | 30,52.68 15 | 30,55.31 16 | 20,47.33 17 | 40,48.74 18 | 20,54.06 19 | 10,42.20 20 | 10,46.23 21 | 50,53.36 22 | 30,53.37 23 | 50,58.52 24 | 40,52.85 25 | 30,49.55 26 | 50,57.34 27 | 10,33.14 28 | 30,50.84 29 | 40,52.24 30 | 20,38.26 31 | 40,52.90 32 | 40,55.97 33 | 20,44.77 34 | 20,56.27 35 | 10,45.16 36 | 30,50.66 37 | 50,51.05 38 | 50,54.77 39 | 10,28.60 40 | 10,39.26 41 | 50,54.61 42 | 40,54.77 43 | 10,48.64 44 | 30,49.12 45 | 20,42.13 46 | 40,58.75 47 | 50,57.70 48 | 20,47.81 49 | 40,55.70 50 | 30,50.15 51 | 50,49.62 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/cifar_aet_k10_sparselabelpropagation_accuracy.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-07-16_17:36 2 | 20,9.40 3 | 10,14.95 4 | 10,13.11 5 | 50,15.00 6 | 40,10.57 7 | 30,10.48 8 | 30,7.64 9 | 40,16.89 10 | 20,12.28 11 | 50,6.80 12 | 10,9.87 13 | 20,9.80 14 | 20,12.39 15 | 30,13.76 16 | 30,13.34 17 | 10,12.35 18 | 40,17.07 19 | 50,8.32 20 | 40,8.52 21 | 50,12.57 22 | 10,10.27 23 | 40,12.18 24 | 20,16.67 25 | 30,16.18 26 | 50,6.81 27 | 30,4.32 28 | 10,10.01 29 | 20,16.42 30 | 40,16.42 31 | 50,12.27 32 | 10,11.49 33 | 50,9.57 34 | 20,10.49 35 | 30,14.54 36 | 40,15.56 37 | 10,8.43 38 | 10,10.90 39 | 30,9.44 40 | 20,12.44 41 | 40,18.57 42 | 50,14.97 43 | 20,10.70 44 | 30,10.91 45 | 10,16.76 46 | 40,17.93 47 | 50,13.42 48 | 20,12.78 49 | 50,10.55 50 | 30,10.50 51 | 40,9.90 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/cifar_aet_k12_poisson_accuracy50trials.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-07-20_14:58 2 | 10,40.80 3 | 20,49.33 4 | 30,52.83 5 | 10,52.69 6 | 50,51.35 7 | 40,53.53 8 | 10,31.75 9 | 20,47.55 10 | 30,53.54 11 | 50,51.99 12 | 20,48.24 13 | 40,52.56 14 | 20,47.18 15 | 30,51.81 16 | 30,55.03 17 | 30,53.25 18 | 10,41.81 19 | 20,53.72 20 | 50,53.18 21 | 10,46.88 22 | 40,48.74 23 | 10,32.91 24 | 40,52.14 25 | 50,58.45 26 | 50,56.86 27 | 40,52.62 28 | 30,50.48 29 | 20,56.10 30 | 30,49.41 31 | 20,38.83 32 | 20,44.95 33 | 40,52.84 34 | 10,44.80 35 | 30,50.29 36 | 50,55.00 37 | 40,55.68 38 | 50,50.86 39 | 40,58.21 40 | 50,57.70 41 | 40,54.54 42 | 10,38.91 43 | 50,54.27 44 | 10,28.73 45 | 20,41.68 46 | 30,49.37 47 | 40,55.61 48 | 30,49.13 49 | 10,48.35 50 | 20,47.40 51 | 50,49.62 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/cifar_aet_k14_poisson_accuracy50trials.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-07-20_14:59 2 | 10,52.50 3 | 10,40.73 4 | 40,53.28 5 | 20,48.99 6 | 30,53.16 7 | 30,52.55 8 | 50,50.65 9 | 20,47.28 10 | 10,31.71 11 | 20,47.58 12 | 50,52.08 13 | 40,52.37 14 | 30,51.64 15 | 30,54.80 16 | 50,53.23 17 | 20,47.38 18 | 40,52.57 19 | 10,41.48 20 | 10,46.40 21 | 30,52.96 22 | 50,56.63 23 | 50,58.70 24 | 40,48.19 25 | 20,38.52 26 | 20,52.78 27 | 10,31.66 28 | 10,45.33 29 | 40,55.73 30 | 30,49.51 31 | 40,51.92 32 | 40,52.97 33 | 20,44.44 34 | 30,49.99 35 | 30,49.75 36 | 20,56.05 37 | 50,54.73 38 | 50,50.72 39 | 30,49.11 40 | 40,53.51 41 | 20,41.63 42 | 10,38.59 43 | 50,57.48 44 | 50,54.32 45 | 40,57.28 46 | 10,48.25 47 | 10,28.55 48 | 20,47.53 49 | 40,55.16 50 | 30,49.69 51 | 50,49.13 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/cifar_aet_k16_poisson_accuracy50trials.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-07-20_15:00 2 | 20,48.75 3 | 10,40.48 4 | 40,53.38 5 | 30,52.60 6 | 20,46.14 7 | 30,52.37 8 | 10,31.63 9 | 50,50.44 10 | 10,52.21 11 | 20,47.01 12 | 50,51.85 13 | 20,46.89 14 | 40,52.56 15 | 50,58.50 16 | 30,54.82 17 | 20,52.63 18 | 50,52.92 19 | 30,51.10 20 | 10,45.74 21 | 40,48.30 22 | 10,41.03 23 | 40,51.66 24 | 10,31.34 25 | 30,52.76 26 | 40,52.37 27 | 50,56.81 28 | 20,55.53 29 | 30,50.17 30 | 30,48.98 31 | 20,38.51 32 | 20,43.93 33 | 40,52.73 34 | 10,45.96 35 | 30,50.09 36 | 40,55.46 37 | 50,54.27 38 | 50,50.81 39 | 40,53.82 40 | 40,56.94 41 | 10,28.51 42 | 50,53.93 43 | 10,38.63 44 | 40,55.03 45 | 30,49.29 46 | 20,41.51 47 | 20,47.65 48 | 30,48.55 49 | 50,57.33 50 | 50,49.35 51 | 10,48.16 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/cifar_aet_k18_poisson_accuracy50trials.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-07-20_15:01 2 | 50,50.38 3 | 40,53.18 4 | 20,47.37 5 | 10,52.52 6 | 30,52.71 7 | 20,48.65 8 | 10,40.23 9 | 30,52.12 10 | 50,51.68 11 | 40,52.39 12 | 30,54.96 13 | 10,31.18 14 | 20,45.79 15 | 20,46.20 16 | 10,46.10 17 | 30,50.93 18 | 50,58.52 19 | 20,51.92 20 | 50,53.27 21 | 10,40.94 22 | 40,48.26 23 | 50,57.23 24 | 30,48.68 25 | 40,52.50 26 | 30,52.69 27 | 20,55.33 28 | 10,30.84 29 | 20,38.02 30 | 30,50.16 31 | 40,51.59 32 | 10,46.02 33 | 40,52.59 34 | 40,55.50 35 | 50,50.77 36 | 50,54.01 37 | 40,54.14 38 | 10,38.50 39 | 30,49.91 40 | 20,44.12 41 | 50,53.63 42 | 20,41.28 43 | 10,28.98 44 | 30,48.82 45 | 20,47.53 46 | 40,56.82 47 | 50,57.22 48 | 30,48.29 49 | 40,54.70 50 | 10,47.92 51 | 50,49.26 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/cifar_aet_k20_poisson_accuracy50trials.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-07-20_15:02 2 | 50,50.06 3 | 30,52.19 4 | 40,52.93 5 | 20,48.51 6 | 30,52.54 7 | 20,45.26 8 | 10,52.42 9 | 10,30.58 10 | 20,47.05 11 | 10,39.81 12 | 50,51.57 13 | 40,52.44 14 | 10,40.90 15 | 30,50.90 16 | 30,54.66 17 | 20,45.97 18 | 50,58.51 19 | 50,53.07 20 | 20,51.33 21 | 40,48.05 22 | 40,52.73 23 | 10,46.04 24 | 10,46.35 25 | 40,52.38 26 | 30,52.60 27 | 50,56.96 28 | 10,30.66 29 | 20,37.05 30 | 30,48.57 31 | 40,51.59 32 | 40,55.35 33 | 30,49.99 34 | 20,55.03 35 | 50,50.62 36 | 50,54.18 37 | 20,41.32 38 | 10,28.40 39 | 40,53.54 40 | 30,49.63 41 | 20,43.95 42 | 10,37.74 43 | 50,53.58 44 | 10,48.02 45 | 20,47.17 46 | 30,48.42 47 | 40,56.47 48 | 40,54.67 49 | 50,57.39 50 | 30,48.64 51 | 50,49.07 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/cifar_aet_k6_poisson_accuracy50trials.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-07-20_14:55 2 | 10,41.01 3 | 20,49.17 4 | 50,52.59 5 | 30,53.11 6 | 10,53.53 7 | 10,29.94 8 | 40,54.45 9 | 40,51.65 10 | 50,52.60 11 | 20,46.77 12 | 10,43.63 13 | 30,54.53 14 | 50,53.60 15 | 30,55.22 16 | 20,47.82 17 | 10,45.89 18 | 30,53.21 19 | 20,48.12 20 | 30,54.77 21 | 50,58.74 22 | 10,34.14 23 | 40,49.47 24 | 20,54.89 25 | 40,53.77 26 | 40,53.76 27 | 10,42.87 28 | 50,57.30 29 | 20,42.36 30 | 30,49.82 31 | 30,49.53 32 | 50,54.64 33 | 10,39.04 34 | 40,52.72 35 | 20,55.03 36 | 40,56.50 37 | 50,55.37 38 | 20,44.10 39 | 30,48.08 40 | 50,50.51 41 | 30,50.92 42 | 40,55.79 43 | 10,29.66 44 | 20,39.18 45 | 30,51.84 46 | 40,58.91 47 | 10,49.53 48 | 20,48.50 49 | 50,58.28 50 | 40,56.06 51 | 50,49.56 52 | -------------------------------------------------------------------------------- /ResultsFromPaper/cifar_aet_k8_poisson_accuracy50trials.csv: -------------------------------------------------------------------------------- 1 | Date/Time, 2020-07-20_14:56 2 | 30,52.79 3 | 50,52.00 4 | 40,53.96 5 | 10,53.19 6 | 50,52.35 7 | 20,47.90 8 | 10,41.07 9 | 10,29.76 10 | 30,54.09 11 | 40,51.52 12 | 20,47.63 13 | 20,49.37 14 | 30,52.85 15 | 10,45.93 16 | 10,44.07 17 | 30,54.95 18 | 20,47.83 19 | 50,53.73 20 | 20,54.54 21 | 50,59.05 22 | 40,49.33 23 | 50,57.05 24 | 30,53.84 25 | 40,52.74 26 | 10,34.33 27 | 40,53.26 28 | 20,55.84 29 | 30,49.88 30 | 40,56.05 31 | 20,39.41 32 | 30,50.59 33 | 40,52.85 34 | 20,44.52 35 | 10,44.32 36 | 40,55.28 37 | 50,50.74 38 | 50,54.69 39 | 30,48.97 40 | 30,50.66 41 | 50,57.71 42 | 10,28.92 43 | 10,40.77 44 | 50,55.12 45 | 40,59.23 46 | 20,41.29 47 | 20,48.20 48 | 40,55.79 49 | 10,49.66 50 | 30,50.72 51 | 50,50.02 52 | -------------------------------------------------------------------------------- /c_code/LICENSE.txt: -------------------------------------------------------------------------------- 1 | Copyright (c) 2014, Laurens van der Maaten (Delft University of Technology) 2 | All rights reserved. 3 | 4 | Redistribution and use in source and binary forms, with or without 5 | modification, are permitted provided that the following conditions are met: 6 | 1. Redistributions of source code must retain the above copyright 7 | notice, this list of conditions and the following disclaimer. 8 | 2. Redistributions in binary form must reproduce the above copyright 9 | notice, this list of conditions and the following disclaimer in the 10 | documentation and/or other materials provided with the distribution. 11 | 3. All advertising materials mentioning features or use of this software 12 | must display the following acknowledgement: 13 | This product includes software developed by the Delft University of Technology. 14 | 4. Neither the name of the Delft University of Technology nor the names of 15 | its contributors may be used to endorse or promote products derived from 16 | this software without specific prior written permission. 17 | 18 | THIS SOFTWARE IS PROVIDED BY LAURENS VAN DER MAATEN ''AS IS'' AND ANY EXPRESS 19 | OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES 20 | OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO 21 | EVENT SHALL LAURENS VAN DER MAATEN BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, 22 | SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, 23 | PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR 24 | BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 25 | CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING 26 | IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY 27 | OF SUCH DAMAGE. 28 | -------------------------------------------------------------------------------- /c_code/hjsolvers.h: -------------------------------------------------------------------------------- 1 | /* dijkstra.h: C code for acceleration of graphlearning library 2 | * 3 | * * Author: Jeff Calder, 2020. 4 | * 5 | */ 6 | 7 | #include 8 | #include 9 | #include "vector_operations.h" 10 | #include "memory_allocation.h" 11 | 12 | 13 | void dijkstra_main(double *d, int *l, int *WI, int *K, double *WV, int *II, double *g, double *f, bool prog, int n, int M, int k, double max_dist); 14 | void dijkstra_hl_main(double *d, int *l, int *WI, int *K, double *WV, int *II, double *g, double *f, bool prog, int n, int M, int k, double max_dist); 15 | void peikonal_main(double *u, int *WI, int *K, double *WV, int *II, double *f, double *g, double p_val, int max_num_it, double tol, int num_bisection_it, bool prog, int n, int M, int k); 16 | void peikonal_fmm_main(double *u, int *WI, int *K, double *WV, int *II, double *f, double *g, double p_val, int num_bisection_it, int n, int M, int k); 17 | -------------------------------------------------------------------------------- /c_code/lp_iterate.h: -------------------------------------------------------------------------------- 1 | /* lplearn.h - C code acceleration for Lplearning 2 | * 3 | * 4 | * Used by the graphlearning package 5 | * 6 | * Inputs: 7 | * uu = initial upper values nx1 array 8 | * ul = initial lower values nx1 array 9 | * (I,J) = indices of adjacency matrix, each nx1 10 | * W = nx1 vector of weights for (I,J) entry 11 | * ind = indices of Dirichlet conditions 12 | * val = values of Dirichlet conditions 13 | * T = number of iterations 14 | * tol = tolerance 15 | * prog = toggles progress indicator 16 | * 17 | * Outputs: 18 | * uu = upper learned function nx1 19 | * ul = lowerf learned function nx1 20 | * r = residual nx1 21 | * 22 | * NOTE: Must supply uu,ul with uu=ul at ind points, uu a supersolution, ul a subsolution, and uu>=ul everywhere 23 | * 24 | * Author: Jeff Calder, 2019. 25 | * 26 | */ 27 | 28 | 29 | #include 30 | #include "vector_operations.h" 31 | #include "memory_allocation.h" 32 | 33 | void lp_iterate_main(double *uu, double *ul, int *II, int *J, double *W, int *ind, double *val, double p, int T, double tol, bool prog, int n, int M, int m); 34 | void lip_iterate_main(double *u, int *II, int *J, double *W, int *ind, double *val, int T, double tol, bool prog, int n, int M, int m, double alpha, double beta); 35 | void lip_iterate_weighted_main(double *u, int *II, int *J, double *W, int *ind, double *val, int T, double tol, bool prog, int n, int M, int m); 36 | 37 | 38 | -------------------------------------------------------------------------------- /c_code/maj_implicit_heap.h: -------------------------------------------------------------------------------- 1 | #include 2 | #ifndef BRANCHING_NUMBER 3 | #define BRANCHING_NUMBER 8 4 | #endif 5 | #ifndef MAJ_IMPLICIT_HEAP 6 | #define MAJ_IMPLICIT_HEAP 7 | 8 | typedef struct{ 9 | key_type key; 10 | int originalIndex; 11 | }implicit_heap_node; 12 | 13 | typedef struct{ 14 | implicit_heap_node *root; 15 | int *locations; 16 | int count; 17 | }implicit_heap; 18 | 19 | 20 | 21 | static void swap_nodes(implicit_heap *heap,int nodeIndex1, int nodeIndex2){ 22 | int i1=heap->root[nodeIndex1].originalIndex; 23 | int i2=heap->root[nodeIndex2].originalIndex; 24 | key_type key1=heap->root[nodeIndex1].key; 25 | key_type key2=heap->root[nodeIndex2].key; 26 | heap->locations[i1]=nodeIndex2; 27 | heap->locations[i2]=nodeIndex1; 28 | heap->root[nodeIndex1].originalIndex=i2; 29 | heap->root[nodeIndex2].originalIndex=i1; 30 | heap->root[nodeIndex1].key=key2; 31 | heap->root[nodeIndex2].key=key1; 32 | } 33 | 34 | static void add_node_to_bottom(implicit_heap *heap, int originalIndex, key_type key){ 35 | 36 | int count=heap->count; 37 | heap->root[count].originalIndex=originalIndex; 38 | heap->root[count].key=key; 39 | heap->locations[originalIndex]=count; 40 | heap->count++; 41 | } 42 | 43 | static void bubble_up(implicit_heap *heap, int nodeIndex){ 44 | 45 | while (nodeIndex>0) { 46 | key_type myKey=heap->root[nodeIndex].key; 47 | int parentIndex=nodeIndex/BRANCHING_NUMBER-((nodeIndex%BRANCHING_NUMBER)==0); 48 | key_type parentKey=heap->root[parentIndex].key; 49 | if(myKeycount; 63 | int childIndex=nodeIndex*BRANCHING_NUMBER+1; 64 | while(childIndexroot[nodeIndex].key; 67 | key_type min=myKey; 68 | for(i=0;iroot[childIndex+i].key; 71 | if(childKeycount; 90 | swap_nodes(heap,count-1,0); 91 | heap->count--; 92 | push_down(heap,0); 93 | } 94 | 95 | 96 | static void decrease_key(implicit_heap *heap, int nodeIndex, key_type newKey){ 97 | 98 | heap->root[nodeIndex].key=newKey; 99 | bubble_up(heap,nodeIndex); 100 | } 101 | 102 | 103 | static void insert_node(implicit_heap *heap, int nodeIndex, key_type newKey){ 104 | add_node_to_bottom(heap,nodeIndex,newKey); 105 | bubble_up(heap,heap->locations[nodeIndex]); 106 | 107 | } 108 | 109 | static implicit_heap create_empty_heap_with_locations(int pcount){ 110 | implicit_heap heap; 111 | heap.count=0; 112 | heap.root=(implicit_heap_node*)malloc(pcount*sizeof(implicit_heap_node)); 113 | heap.locations=(int*)malloc(pcount*sizeof(int)); 114 | return heap; 115 | } 116 | 117 | /*static implicit_heap create_heap_with_batch(key_type *keys, int pcount, int dstride){ 118 | int i; 119 | implicit_heap heap; 120 | heap.root=malloc(pcount*sizeof(implicit_heap_node)); 121 | heap.locations=malloc(pcount*sizeof(int)); 122 | heap.count=0; 123 | 124 | 125 | 126 | for(i=0;icount=0; 141 | } 142 | 143 | 144 | static int empty(implicit_heap *heap){ 145 | return heap->count==0; 146 | } 147 | 148 | static void free_heap(implicit_heap *heap){ 149 | free(heap->locations); 150 | free(heap->root); 151 | } 152 | 153 | #endif 154 | 155 | 156 | 157 | 158 | 159 | 160 | 161 | -------------------------------------------------------------------------------- /c_code/maj_simple_implicit_heap.h: -------------------------------------------------------------------------------- 1 | #include 2 | #ifndef BRANCHING_NUMBER 3 | #define BRANCHING_NUMBER 8 4 | #endif 5 | #ifndef MAJ_SIMPLE_IMPLICIT_HEAP 6 | #define MAJ_SIMPLE_IMPLICIT_HEAP 7 | 8 | typedef struct{ 9 | k_type key; 10 | int originalIndex; 11 | }s_heap_node; 12 | 13 | typedef struct{ 14 | s_heap_node *root; 15 | int count; 16 | }s_heap; 17 | 18 | 19 | 20 | static void s_heap_swap_nodes(s_heap *heap,int nodeIndex1, int nodeIndex2){ 21 | int i1=heap->root[nodeIndex1].originalIndex; 22 | int i2=heap->root[nodeIndex2].originalIndex; 23 | k_type key1=heap->root[nodeIndex1].key; 24 | k_type key2=heap->root[nodeIndex2].key; 25 | heap->root[nodeIndex1].originalIndex=i2; 26 | heap->root[nodeIndex2].originalIndex=i1; 27 | heap->root[nodeIndex1].key=key2; 28 | heap->root[nodeIndex2].key=key1; 29 | } 30 | 31 | static void s_heap_add_node_to_bottom(s_heap *heap, int originalIndex, k_type key){ 32 | 33 | int count=heap->count; 34 | heap->root[count].originalIndex=originalIndex; 35 | heap->root[count].key=key; 36 | heap->count++; 37 | } 38 | 39 | static void s_heap_bubble_up(s_heap *heap, int nodeIndex){ 40 | 41 | while (nodeIndex>0) { 42 | k_type myKey=heap->root[nodeIndex].key; 43 | int parentIndex=nodeIndex/BRANCHING_NUMBER-((nodeIndex%BRANCHING_NUMBER)==0); 44 | k_type parentKey=heap->root[parentIndex].key; 45 | if(myKeycount; 59 | int childIndex=nodeIndex*BRANCHING_NUMBER+1; 60 | while(childIndexroot[nodeIndex].key; 63 | k_type min=myKey; 64 | for(i=0;iroot[childIndex+i].key; 67 | if(childKeycount; 86 | s_heap_swap_nodes(heap,count-1,0); 87 | heap->count--; 88 | s_heap_push_down(heap,0); 89 | } 90 | 91 | 92 | static void s_heap_decrease_key(s_heap *heap, int nodeIndex, k_type newKey){ 93 | 94 | heap->root[nodeIndex].key=newKey; 95 | s_heap_bubble_up(heap,nodeIndex); 96 | }*/ 97 | 98 | 99 | static void s_heap_insert_node(s_heap *heap, int nodeIndex, k_type newKey){ 100 | s_heap_add_node_to_bottom(heap,nodeIndex,newKey); 101 | s_heap_bubble_up(heap,heap->count-1); 102 | 103 | } 104 | 105 | static s_heap s_heap_create_empty_heap(int pcount){ 106 | s_heap heap; 107 | heap.count=0; 108 | heap.root=(s_heap_node*)malloc(pcount*sizeof(s_heap_node)); 109 | return heap; 110 | } 111 | 112 | 113 | /*static int s_heap_empty(s_heap *heap){ 114 | return heap->count==0; 115 | }*/ 116 | 117 | static void s_heap_free_heap(s_heap *heap){ 118 | 119 | free(heap->root); 120 | } 121 | 122 | 123 | #endif 124 | 125 | 126 | 127 | 128 | 129 | 130 | -------------------------------------------------------------------------------- /c_code/mbo_convolution.h: -------------------------------------------------------------------------------- 1 | #include 2 | #include 3 | #include 4 | #include 5 | //#include 6 | #include 7 | 8 | 9 | 10 | 11 | typedef struct{ 12 | 13 | int *neighbors; 14 | float *connectionStrengths; 15 | }knnConvolutionStructure; 16 | 17 | 18 | 19 | 20 | typedef struct{ 21 | int *neighbors; 22 | float *connectionStrengths; 23 | int *counts; 24 | 25 | }generalConvolutionStructure; 26 | 27 | typedef struct{ 28 | int index; 29 | float dist; 30 | }indexedFloat; 31 | 32 | typedef struct{ 33 | float x; 34 | float y; 35 | }mpoint; 36 | 37 | typedef struct{ 38 | int to; 39 | int from; 40 | int index; 41 | 42 | }changedClass; 43 | 44 | typedef struct{ 45 | int index; 46 | int to; 47 | int from; 48 | }nodeChanged; 49 | 50 | typedef struct{ 51 | nodeChanged *updateList; 52 | int *indicies; 53 | float *weights; 54 | int *nncounts; 55 | float *surfaceTensions; 56 | unsigned char *fixedLabels; 57 | int *labels; 58 | float *linear; 59 | int *linearBest; 60 | int *classCounts; 61 | float stoppingCriterion; 62 | float epsilon; 63 | float temperature; 64 | float convexityParameter; 65 | float upperVolumeMultiplier; 66 | float lowerVolumeMultiplier; 67 | int k; 68 | int pcount; 69 | int lcount; 70 | int maxIters; 71 | char singleGrowth; 72 | }mbo_struct; 73 | 74 | void normalize_matrix_with_linear(generalConvolutionStructure g, float *linear, int pcount, int lcount); 75 | 76 | void free_generalConvolutionStructure(generalConvolutionStructure g); 77 | 78 | generalConvolutionStructure create_symmetric_adjacency_matrix(indexedFloat *neighborData, int pcount, int maxNeighbors, int k); 79 | 80 | generalConvolutionStructure create_symmetric_matrix(float (*kernel)(float), indexedFloat *neighborData, int pcount, int maxNeighbors, int k); 81 | void normalize_matrix(generalConvolutionStructure g, int pcount); 82 | 83 | void voronoi_initialization(mbo_struct *mbos, generalConvolutionStructure g, int maxNeighbors, float distanceExponent, int lin); 84 | void bellman_ford_voronoi_initialization(mbo_struct *mbos, generalConvolutionStructure g, float distanceExponent, int lin); 85 | 86 | 87 | void reweight_fidelity_nodes(mbo_struct mbos); 88 | 89 | generalConvolutionStructure create_dual_convolution_structure(mbo_struct mbos); 90 | 91 | float run_mbo(mbo_struct mbos, char mode); 92 | float run_mbo_distance(mbo_struct mbos, float *graphDistances, char mode); 93 | 94 | float run_mbo_efficient(mbo_struct mbos, char mode); 95 | float run_mbo_single_growth_efficient(mbo_struct mbos, char mode); 96 | float run_mbo_with_temperature(mbo_struct mbos, char mode); 97 | float run_mbo_convexity(mbo_struct mbos); 98 | void run_mbo_sgd(mbo_struct mbos); 99 | 100 | void run_k_means_mbo(mbo_struct mbos, float *rawData, int dataDim, char mode); 101 | -------------------------------------------------------------------------------- /c_code/memory_allocation.cpp: -------------------------------------------------------------------------------- 1 | /* memory_allocation.c - 2 | * 3 | * Basic vector and matrix memory allocation 4 | * 5 | * Author: Jeff Calder, 2018. 6 | */ 7 | 8 | #include "stdlib.h" 9 | #include "stdbool.h" 10 | #include "math.h" 11 | #include "memory_allocation.h" 12 | 13 | /*Allocate memory for a mxn array of int and initialize to val*/ 14 | int** array_int(int m, int n, int val){ 15 | 16 | int **ptr = (int**)malloc(m*sizeof(int*)); 17 | ptr[0] = (int*)malloc(m*n*sizeof(int)); 18 | int i,j; 19 | for(i=0;i 3 | #include 4 | #include "memory_allocation.h" 5 | 6 | 7 | //reads a file and returns the bytes 8 | 9 | void *getFileData(const char *fileName){ 10 | 11 | FILE *f=fopen(fileName,"r"); 12 | size_t fsize; 13 | fseek(f,0L,SEEK_END); 14 | fsize=ftell(f); 15 | fseek(f,0L,SEEK_SET); 16 | void *data=malloc(fsize); 17 | size_t o = fread(data,1,fsize,f); 18 | if(o==0){ 19 | printf("Data missing"); 20 | assert(0); 21 | } 22 | fclose(f); 23 | return data; 24 | } 25 | 26 | int compare_int(const void *p, const void *q){ 27 | int i1=*((int *) p); 28 | int i2=*((int *) q); 29 | 30 | if(i2>i1){ 31 | return -1; 32 | }else if(i1>i2){ 33 | return 1; 34 | }else{ 35 | return 0; 36 | } 37 | } 38 | 39 | int compare_float(const void *p, const void *q){ 40 | int i1=*((float *) p); 41 | int i2=*((float *) q); 42 | 43 | if(i2>i1){ 44 | return -1; 45 | }else if(i1>i2){ 46 | return 1; 47 | }else{ 48 | return 0; 49 | } 50 | } 51 | 52 | int compare_indexed_floats(const void *p, const void *q){ 53 | 54 | indexedFloat f1=*((indexedFloat *) p); 55 | indexedFloat f2=*((indexedFloat *) q); 56 | 57 | if(f2.dist>f1.dist){ 58 | return -1; 59 | }else if(f1.dist>f2.dist){ 60 | return 1; 61 | }else{ 62 | return 0; 63 | } 64 | 65 | } 66 | 67 | void sort_indexed_floats(indexedFloat *distances, int num){ 68 | 69 | qsort(distances, num, sizeof(indexedFloat), &compare_indexed_floats); 70 | 71 | } 72 | void sort_int(int *array, int n){ 73 | qsort(array,n,sizeof(int),&compare_int); 74 | } 75 | void sort_float(float *array, int n){ 76 | qsort(array,n,sizeof(float),&compare_float); 77 | } 78 | /* 79 | indexedFloat *readAndConvertDistanceDataToIndexedFloat(const char *distanceFilename, const char *knnFilename, int maxNeighbors, int pcount){ 80 | int i,j; 81 | float *distances=getFileData(distanceFilename); 82 | int *indicies=getFileData(knnFilename); 83 | indexedFloat *comboData=calloc(maxNeighbors*pcount,sizeof(indexedFloat)); 84 | for(i=0;i0){ 211 | run_mbo_with_temperature(mbos, mode[0]); 212 | }else{ 213 | run_mbo(mbos, mode[0]); 214 | } 215 | 216 | 217 | } 218 | 219 | 220 | 221 | 222 | -------------------------------------------------------------------------------- /c_code/mnist_benchmark.h: -------------------------------------------------------------------------------- 1 | #include "mbo_convolution.h" 2 | #include 3 | #include 4 | #include 5 | 6 | 7 | void mbo_main(int *labels, int *II, int *J, float *W, int *ind, int *val, int *classCounts, bool prog, int pcount, int M, int m, int lcount, int maxIters, float stoppingCriterion, float temperature, float upperVolumeMultiplier, float lowerVolumeMultiplier); 8 | 9 | void mbo_main_original(int k, int numTrials, float trainingFraction, int maxIters, float StoppingCriterion, float temperature, float upperVolumeMultiplier, float lowerVolumeMultiplier); 10 | -------------------------------------------------------------------------------- /c_code/sptree.h: -------------------------------------------------------------------------------- 1 | /* 2 | * 3 | * Copyright (c) 2014, Laurens van der Maaten (Delft University of Technology) 4 | * All rights reserved. 5 | * 6 | * Redistribution and use in source and binary forms, with or without 7 | * modification, are permitted provided that the following conditions are met: 8 | * 1. Redistributions of source code must retain the above copyright 9 | * notice, this list of conditions and the following disclaimer. 10 | * 2. Redistributions in binary form must reproduce the above copyright 11 | * notice, this list of conditions and the following disclaimer in the 12 | * documentation and/or other materials provided with the distribution. 13 | * 3. All advertising materials mentioning features or use of this software 14 | * must display the following acknowledgement: 15 | * This product includes software developed by the Delft University of Technology. 16 | * 4. Neither the name of the Delft University of Technology nor the names of 17 | * its contributors may be used to endorse or promote products derived from 18 | * this software without specific prior written permission. 19 | * 20 | * THIS SOFTWARE IS PROVIDED BY LAURENS VAN DER MAATEN ''AS IS'' AND ANY EXPRESS 21 | * OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES 22 | * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO 23 | * EVENT SHALL LAURENS VAN DER MAATEN BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, 24 | * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, 25 | * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR 26 | * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 27 | * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING 28 | * IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY 29 | * OF SUCH DAMAGE. 30 | * 31 | */ 32 | 33 | 34 | #ifndef SPTREE_H 35 | #define SPTREE_H 36 | 37 | using namespace std; 38 | 39 | 40 | class Cell { 41 | 42 | unsigned int dimension; 43 | double* corner; 44 | double* width; 45 | 46 | 47 | public: 48 | Cell(unsigned int inp_dimension); 49 | Cell(unsigned int inp_dimension, double* inp_corner, double* inp_width); 50 | ~Cell(); 51 | 52 | double getCorner(unsigned int d); 53 | double getWidth(unsigned int d); 54 | void setCorner(unsigned int d, double val); 55 | void setWidth(unsigned int d, double val); 56 | bool containsPoint(double point[]); 57 | }; 58 | 59 | 60 | class SPTree 61 | { 62 | 63 | // Fixed constants 64 | static const unsigned int QT_NODE_CAPACITY = 1; 65 | 66 | // A buffer we use when doing force computations 67 | double* buff; 68 | 69 | // Properties of this node in the tree 70 | SPTree* parent; 71 | unsigned int dimension; 72 | bool is_leaf; 73 | unsigned int size; 74 | unsigned int cum_size; 75 | 76 | // Axis-aligned bounding box stored as a center with half-dimensions to represent the boundaries of this quad tree 77 | Cell* boundary; 78 | 79 | // Indices in this space-partitioning tree node, corresponding center-of-mass, and list of all children 80 | double* data; 81 | double* center_of_mass; 82 | unsigned int index[QT_NODE_CAPACITY]; 83 | 84 | // Children 85 | SPTree** children; 86 | unsigned int no_children; 87 | 88 | public: 89 | SPTree(unsigned int D, double* inp_data, unsigned int N); 90 | SPTree(unsigned int D, double* inp_data, double* inp_corner, double* inp_width); 91 | SPTree(unsigned int D, double* inp_data, unsigned int N, double* inp_corner, double* inp_width); 92 | SPTree(SPTree* inp_parent, unsigned int D, double* inp_data, unsigned int N, double* inp_corner, double* inp_width); 93 | SPTree(SPTree* inp_parent, unsigned int D, double* inp_data, double* inp_corner, double* inp_width); 94 | ~SPTree(); 95 | void setData(double* inp_data); 96 | SPTree* getParent(); 97 | void construct(Cell boundary); 98 | bool insert(unsigned int new_index); 99 | void subdivide(); 100 | bool isCorrect(); 101 | void rebuildTree(); 102 | void getAllIndices(unsigned int* indices); 103 | unsigned int getDepth(); 104 | void computeNonEdgeForces(unsigned int point_index, double theta, double neg_f[], double* sum_Q, double theta2); 105 | void computeEdgeForces(unsigned int* row_P, unsigned int* col_P, double* val_P, int N, double* pos_f, double theta1); 106 | void print(); 107 | 108 | private: 109 | void init(SPTree* inp_parent, unsigned int D, double* inp_data, double* inp_corner, double* inp_width); 110 | void fill(unsigned int N); 111 | unsigned int getAllIndices(unsigned int* indices, unsigned int loc); 112 | bool isChild(unsigned int test_index, unsigned int start, unsigned int end); 113 | }; 114 | 115 | #endif 116 | -------------------------------------------------------------------------------- /c_code/tsne.h: -------------------------------------------------------------------------------- 1 | /* 2 | * 3 | * Copyright (c) 2014, Laurens van der Maaten (Delft University of Technology) 4 | * All rights reserved. 5 | * 6 | * Redistribution and use in source and binary forms, with or without 7 | * modification, are permitted provided that the following conditions are met: 8 | * 1. Redistributions of source code must retain the above copyright 9 | * notice, this list of conditions and the following disclaimer. 10 | * 2. Redistributions in binary form must reproduce the above copyright 11 | * notice, this list of conditions and the following disclaimer in the 12 | * documentation and/or other materials provided with the distribution. 13 | * 3. All advertising materials mentioning features or use of this software 14 | * must display the following acknowledgement: 15 | * This product includes software developed by the Delft University of Technology. 16 | * 4. Neither the name of the Delft University of Technology nor the names of 17 | * its contributors may be used to endorse or promote products derived from 18 | * this software without specific prior written permission. 19 | * 20 | * THIS SOFTWARE IS PROVIDED BY LAURENS VAN DER MAATEN ''AS IS'' AND ANY EXPRESS 21 | * OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES 22 | * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO 23 | * EVENT SHALL LAURENS VAN DER MAATEN BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, 24 | * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, 25 | * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR 26 | * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN 27 | * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING 28 | * IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY 29 | * OF SUCH DAMAGE. 30 | * 31 | */ 32 | 33 | 34 | //#ifndef TSNE_H 35 | //#define TSNE_H 36 | 37 | //#ifdef __cplusplus 38 | //extern "C" { 39 | //namespace TSNE { 40 | //#endif 41 | // void run(double* X, int N, int D, double* Y, int no_dims, double perplexity, double theta, int rand_seed, bool skip_random_init, int max_iter, int stop_lying_iter, int mom_switch_iter); 42 | void tsne_run(double* X, int N, int D, double* Y, int no_dims, double perplexity, double theta, int rand_seed, bool skip_random_init, int max_iter, int stop_lying_iter, int mom_switch_iter, double time_step, double theta1, double theta2, double alpha, int num_early, bool prog, bool dump_to_file); 43 | // bool load_data(double** data, int* n, int* d, int* no_dims, double* theta, double* perplexity, int* rand_seed, int* max_iter); 44 | bool load_data(double** data, int* n, int* d, int* no_dims, double* theta, double* perplexity, int* rand_seed, int* max_iter, double *time_step, double *theta1, double *theta2, double *alpha, int *num_early); 45 | void save_data(double* data, int* landmarks, double* costs, int n, int d); 46 | //#ifdef __cplusplus 47 | //}; 48 | //} 49 | //#endif 50 | 51 | //#endif 52 | -------------------------------------------------------------------------------- /c_code/vector_operations.h: -------------------------------------------------------------------------------- 1 | /* vector_operations.h - 2 | * 3 | * Basic vector calculus operations. 4 | * 5 | * Author: Jeff Calder, 2018. 6 | */ 7 | 8 | #define ABS(a) (((a)<0)?-(a):(a)) 9 | #define SIGN(a) (((a)<0)?(-1):(1)) 10 | #define MIN(a,b) (((a)<(b))?(a):(b)) 11 | #define MIN3(a,b,c) (MIN(MIN(a,b),c)) 12 | #define MAX(a,b) (((a)>(b))?(a):(b)) 13 | #define MAX3(a,b,c) (MAX(MAX(a,b),c)) 14 | #define PI 3.14159265359 15 | 16 | #define dot(x,y) (x[0]*y[0] + x[1]*y[1] + x[2]*y[2]) 17 | #define norm(x) (sqrt(x[0]*x[0] + x[1]*x[1] + x[2]*x[2])) 18 | #define norm_squared(x) (x[0]*x[0] + x[1]*x[1] + x[2]*x[2]) 19 | #define dist(x,y) (sqrt((x[0]-y[0])*(x[0]-y[0]) + (x[1]-y[1])*(x[1]-y[1]) + (x[2]-y[2])*(x[2]-y[2]))) 20 | #define dist_squared(x,y) ((x[0]-y[0])*(x[0]-y[0]) + (x[1]-y[1])*(x[1]-y[1]) + (x[2]-y[2])*(x[2]-y[2])) 21 | #define cross(x,y,z) z[0] = x[1]*y[2] - x[2]*y[1]; z[1] = x[2]*y[0] - x[0]*y[2]; z[2] = x[0]*y[1] - x[1]*y[0] 22 | #define centroid(x,y,z,p) p[0] = (x[0] + y[0] + z[0])/3; p[1] = (x[1] + y[1] + z[1])/3; p[2] = (x[2] + y[2] + z[2])/3 23 | #define average(x,y,z) z[0] = (x[0] + y[0])/2; z[1] = (x[1] + y[1])/2; z[2] = (x[2] + y[2])/2 24 | #define add(x,y,z) z[0] = x[0] + y[0]; z[1] = x[1] + y[1]; z[2] = x[2] + y[2] 25 | #define sub(x,y,z) z[0] = x[0] - y[0]; z[1] = x[1] - y[1]; z[2] = x[2] - y[2] 26 | #define mult(x,a,z) z[0] = a*x[0]; z[1] = a*x[1]; z[2] = a*x[2] 27 | #define new_coordinates(x,a,b,c) v1 = dot(x,e1); v2 = dot(x,e2); v3 = dot(x,e3);x[0] = v1; x[1] = v2; x[2] = v3 28 | 29 | -------------------------------------------------------------------------------- /docs/_config.yml: -------------------------------------------------------------------------------- 1 | theme: jekyll-theme-slate -------------------------------------------------------------------------------- /examples/RP1D_mnist.py: -------------------------------------------------------------------------------- 1 | import graphlearning as gl 2 | 3 | data, labels = gl.datasets.load('mnist') 4 | 5 | x = data[labels <= 1] 6 | y = labels[labels <= 1] 7 | y_pred = gl.clustering.RP1D(x,20) 8 | 9 | accuracy = gl.clustering.clustering_accuracy(y_pred, y) 10 | print('Clustering Accuracy: %.2f%%'%accuracy) 11 | -------------------------------------------------------------------------------- /examples/al_test.py: -------------------------------------------------------------------------------- 1 | import graphlearning.active_learning as al 2 | import graphlearning as gl 3 | import numpy as np 4 | import matplotlib.pyplot as plt 5 | import sklearn.datasets as datasets 6 | 7 | X,labels = datasets.make_moons(n_samples=500,noise=0.1) 8 | W = gl.weightmatrix.knn(X,10) 9 | train_ind = gl.trainsets.generate(labels, rate=5) 10 | #plt.scatter(X[:,0],X[:,1], c=labels) 11 | #plt.scatter(X[train_ind,0],X[train_ind,1], c='r') 12 | #plt.show() 13 | 14 | model = gl.ssl.laplace(W) 15 | acq = al.model_change_vopt() 16 | act = al.active_learning(W, np.arange(labels.size), train_ind, labels[train_ind], 200) 17 | 18 | for i in range(10): 19 | u = model.fit(act.current_labeled_set, act.current_labels) # perform classification with GSSL classifier 20 | query_points = act.select_query_points(u, acq, oracle=None) # return this iteration's newly chosen points 21 | query_labels = labels[query_points] # simulate the human in the loop process 22 | act.update_labeled_data(query_points, query_labels) # update the active_learning object's labeled set 23 | 24 | # plot 25 | plt.scatter(X[:,0],X[:,1], c=labels) 26 | plt.scatter(X[act.current_labeled_set,0],X[act.current_labeled_set,1], c='r') 27 | plt.scatter(X[query_points,0],X[query_points,1], c='r', marker='*', s=200, edgecolors='k', linewidths=1.5) 28 | plt.show() 29 | print(act.current_labeled_set) 30 | print(act.current_labels) 31 | 32 | -------------------------------------------------------------------------------- /examples/ars_tsne.py: -------------------------------------------------------------------------------- 1 | import graphlearning as gl 2 | import numpy as np 3 | import matplotlib.pyplot as plt 4 | 5 | #Load the MNIST data 6 | data,labels = gl.datasets.load('mnist') 7 | 8 | #In order to run the code more quickly, 9 | #you may want to subsample MNIST. 10 | size = 70000 11 | if size < data.shape[0]: #If less than 70000 12 | ind = np.random.choice(data.shape[0], size=size, replace=False) 13 | data = data[ind,:] 14 | labels = labels[ind] 15 | 16 | #Run ARS t-SNE and plot the result 17 | Y = gl.graph.ars(data, prog=True) 18 | plt.scatter(Y[:,0],Y[:,1],c=labels,s=1) 19 | plt.show() 20 | 21 | -------------------------------------------------------------------------------- /examples/cone.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/examples/cone.png -------------------------------------------------------------------------------- /examples/dijkstra.py: -------------------------------------------------------------------------------- 1 | import graphlearning as gl 2 | import numpy as np 3 | 4 | for n in [int(10**i) for i in range(3,6)]: 5 | 6 | X = np.random.rand(n,2) 7 | X[0,:]=[0.5,0.5] 8 | W = gl.weightmatrix.knn(X,50,kernel='distance') 9 | G = gl.graph(W) 10 | u = G.dijkstra([0]) 11 | 12 | u_true = np.linalg.norm(X - [0.5,0.5],axis=1) 13 | error = np.linalg.norm(u-u_true, ord=np.inf) 14 | print('n = %d, Error = %f'%(n,error)) 15 | 16 | -------------------------------------------------------------------------------- /examples/dijkstra_hl.py: -------------------------------------------------------------------------------- 1 | import graphlearning as gl 2 | import matplotlib.pyplot as plt 3 | import numpy as np 4 | #from mayavi import mlab 5 | 6 | 7 | n = 10000 8 | X = gl.utils.rand_ball(n,2) 9 | X[0,:]=[0,0] 10 | W = gl.weightmatrix.knn(X,50,kernel='distance',symmetrize=True) 11 | G = gl.graph(W) 12 | u = G.dijkstra_hl([0]) 13 | 14 | #Check the residual 15 | grad = G.gradient(u**2,p=-1) 16 | H = grad.max(axis=0).toarray().flatten() 17 | print('Residual=%f'%np.max(np.absolute(H-u))) 18 | 19 | x,y = X[:,0],X[:,1] 20 | plt.scatter(x,y,c=u) 21 | plt.show() 22 | #Tri = gl.utils.mesh(X) 23 | #mlab.figure(size=(1000,1000),bgcolor=(1,1,1)) 24 | #mlab.triangular_mesh(x,y,3*(np.max(u)-u),Tri) 25 | #mlab.savefig('cone.png') 26 | #mlab.show() 27 | #plt.show() 28 | -------------------------------------------------------------------------------- /examples/fokker_planck_clustering.py: -------------------------------------------------------------------------------- 1 | import numpy as np 2 | import graphlearning as gl 3 | import matplotlib.pyplot as plt 4 | 5 | X,L = gl.datasets.two_skies(1000) 6 | W = gl.weightmatrix.knn(X,10) 7 | 8 | knn_ind,knn_dist = gl.weightmatrix.knnsearch(X,50) 9 | rho = 1/np.max(knn_dist,axis=1) 10 | 11 | model = gl.clustering.fokker_planck(W,num_clusters=2,t=1000,beta=0.5,rho=rho) 12 | labels = model.fit_predict() 13 | 14 | plt.scatter(X[:,0],X[:,1], c=labels) 15 | plt.show() 16 | 17 | -------------------------------------------------------------------------------- /examples/incres_mnist.py: -------------------------------------------------------------------------------- 1 | import graphlearning as gl 2 | 3 | W = gl.weightmatrix.knn('mnist', 10, metric='vae') 4 | labels = gl.datasets.load('mnist', labels_only=True) 5 | 6 | model = gl.clustering.incres(W, num_clusters=10) 7 | pred_labels = model.fit_predict(all_labels=labels) 8 | 9 | accuracy = gl.clustering.clustering_accuracy(pred_labels,labels) 10 | print('Clustering Accuracy: %.2f%%'%accuracy) 11 | 12 | -------------------------------------------------------------------------------- /examples/peikonal.py: -------------------------------------------------------------------------------- 1 | import graphlearning as gl 2 | import numpy as np 3 | import matplotlib.pyplot as plt 4 | import time 5 | 6 | 7 | X = np.random.rand(int(10000),2) 8 | x,y = X[:,0],X[:,1] 9 | 10 | eps = 0.025 11 | W = gl.weightmatrix.epsilon_ball(X, eps) 12 | G = gl.graph(W) 13 | 14 | bdy_set = (x < eps) | (x > 1-eps) | (y < eps) | (y > 1-eps) 15 | u = G.peikonal(bdy_set) 16 | 17 | plt.scatter(x,y,c=u,s=0.25) 18 | plt.scatter(x[bdy_set],y[bdy_set],c='r',s=0.5) 19 | plt.show() 20 | -------------------------------------------------------------------------------- /examples/plaplace.py: -------------------------------------------------------------------------------- 1 | import graphlearning as gl 2 | import numpy as np 3 | import matplotlib.pyplot as plt 4 | 5 | X = np.random.rand(int(1e4),2) 6 | x,y = X[:,0],X[:,1] 7 | 8 | eps = 0.02 9 | W = gl.weightmatrix.epsilon_ball(X, eps) 10 | G = gl.graph(W) 11 | 12 | bdy_set = (x < eps) | (x > 1-eps) | (y < eps) | (y > 1-eps) 13 | bdy_val = (x-0.5)**2 + (y-0.5)**2 14 | 15 | u = G.plaplace(bdy_set, bdy_val[bdy_set], p=10) 16 | 17 | plt.scatter(x,y,c=u,s=0.25) 18 | plt.scatter(x[bdy_set],y[bdy_set],c='r',s=0.5) 19 | plt.show() 20 | -------------------------------------------------------------------------------- /examples/poisson_directed.py: -------------------------------------------------------------------------------- 1 | import numpy as np 2 | import graphlearning as gl 3 | import matplotlib.pyplot as plt 4 | import sklearn.datasets as datasets 5 | 6 | X,labels = datasets.make_moons(n_samples=500,noise=0.1) 7 | W = gl.weightmatrix.knn(X,10,symmetrize=False) 8 | 9 | train_ind = gl.trainsets.generate(labels, rate=5) 10 | train_labels = labels[train_ind] 11 | 12 | model = gl.ssl.poisson(W, solver='gradient_descent') 13 | pred_labels = model.fit_predict(train_ind, train_labels) 14 | 15 | accuracy = gl.ssl.ssl_accuracy(pred_labels, labels, train_ind) 16 | print("Accuracy: %.2f%%"%accuracy) 17 | 18 | plt.scatter(X[:,0],X[:,1], c=pred_labels) 19 | plt.scatter(X[train_ind,0],X[train_ind,1], c='r') 20 | plt.show() 21 | 22 | -------------------------------------------------------------------------------- /examples/poisson_mbo.py: -------------------------------------------------------------------------------- 1 | import graphlearning as gl 2 | 3 | labels = gl.datasets.load('mnist', labels_only=True) 4 | W = gl.weightmatrix.knn('mnist', 10, metric='vae') 5 | 6 | num_train_per_class = 1 7 | train_ind = gl.trainsets.generate(labels, rate=num_train_per_class) 8 | train_labels = labels[train_ind] 9 | 10 | class_priors = gl.utils.class_priors(labels) 11 | model = gl.ssl.poisson_mbo(W, class_priors) 12 | pred_labels = model.fit_predict(train_ind,train_labels,all_labels=labels) 13 | 14 | accuracy = gl.ssl.ssl_accuracy(labels,pred_labels,train_ind) 15 | print(model.name + ': %.2f%%'%accuracy) 16 | -------------------------------------------------------------------------------- /examples/randomized_svd.py: -------------------------------------------------------------------------------- 1 | import numpy as np 2 | import matplotlib.pyplot as plt 3 | import sklearn.datasets as datasets 4 | import graphlearning as gl 5 | 6 | X,L = datasets.make_moons(n_samples=500,noise=0.1) 7 | W = gl.weightmatrix.knn(X,10) 8 | G = gl.graph(W) 9 | 10 | num_eig = 7 11 | vals_exact, vecs_exact = G.eigen_decomp(normalization='normalized', k=num_eig, method='exact') 12 | vals_rsvd, vecs_rsvd = G.eigen_decomp(normalization='normalized', k=num_eig, method='lowrank', q=50, c=50) 13 | 14 | for i in range(1,num_eig): 15 | rsvd = vecs_rsvd[:,i] 16 | exact = vecs_exact[:,i] 17 | 18 | sign = np.sum(rsvd*exact) 19 | if sign < 0: 20 | rsvd *= -1 21 | 22 | err = np.max(np.absolute(rsvd - exact))/max(np.max(np.absolute(rsvd)),np.max(np.absolute(exact))) 23 | 24 | fig, (ax1,ax2) = plt.subplots(1,2, figsize=(10,5)) 25 | fig.suptitle('Eigenvector %d, err=%f'%(i,err)) 26 | 27 | ax1.scatter(X[:,0],X[:,1], c=rsvd) 28 | ax1.set_title('Random SVD') 29 | 30 | ax2.scatter(X[:,0],X[:,1], c=exact) 31 | ax2.set_title('Exact') 32 | 33 | plt.show() 34 | -------------------------------------------------------------------------------- /examples/regression.py: -------------------------------------------------------------------------------- 1 | #Graph Laplacian Based Regression: This code shows how to use 2 | #the GraphLearning package to perform graph-Laplacian-based regression 3 | #using the ordinary Laplacian and p-Laplacian. 4 | import numpy as np 5 | from scipy import sparse 6 | import graphlearning as gl 7 | 8 | n=1000 #Number of data points 9 | m=40 #Number of features 10 | v=0.75 #Train set size 11 | lam=0.1 #Ridge regression parameter (larger values encourage more smoothness) 12 | k=20 #Number of neighbors to use in graph 13 | p=5 #p-Laplace value 14 | 15 | #Data, regression function, and train indices/mask 16 | X = np.random.rand(n,m) 17 | y = np.sum(X,axis=1) #Regression function to predict 18 | train_ind = np.random.choice(n,size=int(v*n),replace=False) 19 | train_mask = np.zeros(n,dtype=bool) 20 | train_mask[train_ind]=True 21 | test_mask = ~train_mask 22 | 23 | #Graph-Laplace based regression 24 | #yhat = (B + \lambda L)^{-1} 25 | #yhat = argmin_u \{ || B(u - y) ||^2 + \lambda u^TLu \} 26 | #Diagonal matrix B indicates the locations of training labels. 27 | B = sparse.spdiags(train_mask[None,:].astype(float),0) 28 | W = gl.weightmatrix.knn(X,k) 29 | G = gl.graph(W) 30 | L = G.laplacian() 31 | yhat = gl.utils.conjgrad(B + lam*L,B*y) 32 | 33 | #The commented out code below implements p-Laplacian regression with p given above (p=5) 34 | #yhat = G.plaplace(train_ind,y[train_ind],p) 35 | 36 | #Compare mean squarad error on test set 37 | rmse = np.sqrt(np.mean((yhat[test_mask] - y[test_mask])**2)) 38 | print('RMSE',rmse) 39 | print('Relative RMSE: %.2f%%'%(100*rmse/np.sqrt(np.mean(y**2)))) 40 | -------------------------------------------------------------------------------- /examples/spectral_mnist.py: -------------------------------------------------------------------------------- 1 | import graphlearning as gl 2 | 3 | W = gl.weightmatrix.knn('mnist', 10, metric='vae') 4 | labels = gl.datasets.load('mnist', labels_only=True) 5 | 6 | model = gl.clustering.spectral(W, num_clusters=10, extra_dim=4) 7 | pred_labels = model.fit_predict(all_labels=labels) 8 | 9 | accuracy = gl.clustering.clustering_accuracy(pred_labels,labels) 10 | print('Clustering Accuracy: %.2f%%'%accuracy) 11 | 12 | -------------------------------------------------------------------------------- /examples/spectral_twomoons.py: -------------------------------------------------------------------------------- 1 | import numpy as np 2 | import graphlearning as gl 3 | import matplotlib.pyplot as plt 4 | import sklearn.datasets as datasets 5 | 6 | X,labels = datasets.make_moons(n_samples=500,noise=0.1) 7 | W = gl.weightmatrix.knn(X,10) 8 | 9 | model = gl.clustering.spectral(W, num_clusters=2) 10 | pred_labels = model.fit_predict() 11 | 12 | accuracy = gl.clustering.clustering_accuracy(pred_labels, labels) 13 | print('Clustering Accuracy: %.2f%%'%accuracy) 14 | 15 | plt.scatter(X[:,0],X[:,1], c=pred_labels) 16 | plt.axis('off') 17 | plt.show() 18 | 19 | -------------------------------------------------------------------------------- /examples/ssl_classpriors.py: -------------------------------------------------------------------------------- 1 | import graphlearning as gl 2 | 3 | labels = gl.datasets.load('mnist', labels_only=True) 4 | W = gl.weightmatrix.knn('mnist', 10, metric='vae') 5 | 6 | num_train_per_class = 1 7 | train_ind = gl.trainsets.generate(labels, rate=num_train_per_class) 8 | train_labels = labels[train_ind] 9 | 10 | class_priors = gl.utils.class_priors(labels) 11 | model = gl.ssl.laplace(W, class_priors=class_priors) 12 | model.fit(train_ind,train_labels) 13 | 14 | pred_labels = model.predict(ignore_class_priors=True) 15 | accuracy = gl.ssl.ssl_accuracy(labels,pred_labels,train_ind) 16 | print(model.name + ' without class priors: %.2f%%'%accuracy) 17 | 18 | pred_labels = model.predict() 19 | accuracy = gl.ssl.ssl_accuracy(labels,pred_labels,train_ind) 20 | print(model.name + ' with class priors: %.2f%%'%accuracy) 21 | 22 | 23 | -------------------------------------------------------------------------------- /examples/ssl_mnist.py: -------------------------------------------------------------------------------- 1 | import graphlearning as gl 2 | 3 | labels = gl.datasets.load('mnist', labels_only=True) 4 | 5 | num_train_per_class = 1 6 | train_ind = gl.trainsets.generate(labels, rate=num_train_per_class) 7 | train_labels = labels[train_ind] 8 | 9 | W = gl.weightmatrix.knn('mnist', 10, metric='vae',kernel='gaussian') 10 | models = [gl.ssl.laplace(W), gl.ssl.poisson(W),gl.ssl.plaplace(W,p=3),gl.ssl.amle(W),gl.ssl.volume_mbo(W,gl.utils.class_priors(labels))] 11 | 12 | for model in models: 13 | pred_labels = model.fit_predict(train_ind,train_labels) 14 | accuracy = gl.ssl.ssl_accuracy(labels,pred_labels,train_ind) 15 | print(model.name + ': %.2f%%'%accuracy) 16 | 17 | -------------------------------------------------------------------------------- /examples/ssl_trials.py: -------------------------------------------------------------------------------- 1 | import graphlearning as gl 2 | 3 | dataset = 'mnist' 4 | metric = 'vae' 5 | k = 10 6 | 7 | W = gl.weightmatrix.knn(dataset, k, metric=metric) 8 | D = gl.weightmatrix.knn(dataset, k, metric=metric, kernel='distance') 9 | 10 | labels = gl.datasets.load(dataset, metric=metric, labels_only=True) 11 | trainsets = gl.trainsets.load(dataset) 12 | 13 | model_list = [gl.ssl.graph_nearest_neighbor(D), 14 | gl.ssl.laplace(W), 15 | gl.ssl.laplace(W, reweighting='wnll'), 16 | gl.ssl.laplace(W, reweighting='poisson'), 17 | gl.ssl.poisson(W, solver='gradient_descent')] 18 | 19 | tag = dataset + '_' + metric + '_k%d'%k 20 | for model in model_list: 21 | model.ssl_trials(trainsets, labels, num_cores=20, tag=tag) 22 | 23 | gl.ssl.accuracy_table(model_list, tag=tag, savefile='SSL_'+dataset+'.tex', title="SSL Comparison: "+dataset) 24 | gl.ssl.accuracy_plot(model_list, tag=tag, title='SSL') 25 | -------------------------------------------------------------------------------- /examples/ssl_twomoons.py: -------------------------------------------------------------------------------- 1 | import numpy as np 2 | import graphlearning as gl 3 | import matplotlib.pyplot as plt 4 | import sklearn.datasets as datasets 5 | 6 | X,labels = datasets.make_moons(n_samples=500,noise=0.1) 7 | W = gl.weightmatrix.knn(X,10) 8 | 9 | train_ind = gl.trainsets.generate(labels, rate=5) 10 | train_labels = labels[train_ind] 11 | 12 | model = gl.ssl.laplace(W) 13 | pred_labels = model.fit_predict(train_ind, train_labels) 14 | 15 | accuracy = gl.ssl.ssl_accuracy(pred_labels, labels, train_ind) 16 | print("Accuracy: %.2f%%"%accuracy) 17 | 18 | plt.scatter(X[:,0],X[:,1], c=pred_labels) 19 | plt.scatter(X[train_ind,0],X[train_ind,1], c='r') 20 | plt.show() 21 | 22 | -------------------------------------------------------------------------------- /examples/vae_mnist.py: -------------------------------------------------------------------------------- 1 | import graphlearning as gl 2 | 3 | data, labels = gl.datasets.load('mnist') 4 | data_vae = gl.weightmatrix.vae(data) 5 | 6 | W_raw = gl.weightmatrix.knn(data, 10) 7 | W_vae = gl.weightmatrix.knn(data_vae, 10) 8 | 9 | num_train_per_class = 1 10 | train_ind = gl.trainsets.generate(labels, rate=num_train_per_class) 11 | train_labels = labels[train_ind] 12 | 13 | pred_labels_raw = gl.ssl.poisson(W_raw).fit_predict(train_ind,train_labels) 14 | pred_labels_vae = gl.ssl.poisson(W_vae).fit_predict(train_ind,train_labels) 15 | 16 | accuracy_raw = gl.ssl.ssl_accuracy(labels,pred_labels_raw,train_ind) 17 | accuracy_vae = gl.ssl.ssl_accuracy(labels,pred_labels_vae,train_ind) 18 | 19 | print('Raw Accuracy: %.2f%%'%accuracy_raw) 20 | print('VAE Accuracy: %.2f%%'%accuracy_vae) 21 | -------------------------------------------------------------------------------- /generate_docs.sh: -------------------------------------------------------------------------------- 1 | #!/bin/bash 2 | 3 | if [[ "$OSTYPE" == "darwin"* ]]; then 4 | sed -i '' 's/import sklearn.cluster as cluster/#import sklearn.cluster as cluster/' graphlearning/clustering.py 5 | else 6 | sed -i 's/import sklearn.cluster as cluster/#import sklearn.cluster as cluster/' graphlearning/clustering.py 7 | fi 8 | 9 | pdoc3 --template-dir ./pdoc/templates --html --force -o docs/ ./graphlearning > /dev/null 10 | 11 | if [[ "$OSTYPE" == "darwin"* ]]; then 12 | sed -i '' 's/#import sklearn.cluster as cluster/import sklearn.cluster as cluster/' graphlearning/clustering.py 13 | else 14 | sed -i 's/#import sklearn.cluster as cluster/import sklearn.cluster as cluster/' graphlearning/clustering.py 15 | fi 16 | 17 | mv docs/graphlearning/* docs/ 18 | rmdir docs/graphlearning/ 19 | ls -1 docs/ | sed 's/^/docs\//' 20 | 21 | #Also, it is possible to omit functions by using the module dictionary within pdoc. An easy way to do it is to put the following at the top of each .py file in which you want to omit certain functions: 22 | 23 | #__pdoc__ = {} 24 | #__pdoc__['function_name'] = False 25 | 26 | #If you want to omit class methods instead, simply put 27 | 28 | #__pdoc__ = {} 29 | #__pdoc__['class_name.method_name'] = False 30 | # 31 | #within the class definition. 32 | 33 | -------------------------------------------------------------------------------- /graphlearning/__init__.py: -------------------------------------------------------------------------------- 1 | from graphlearning import utils 2 | from graphlearning import ssl 3 | from graphlearning import weightmatrix 4 | from graphlearning import datasets 5 | from graphlearning import trainsets 6 | from graphlearning import clustering 7 | from graphlearning import active_learning 8 | from graphlearning.graph import graph 9 | 10 | -------------------------------------------------------------------------------- /graphlearning/trainsets.py: -------------------------------------------------------------------------------- 1 | """ 2 | Trainsets 3 | ========== 4 | 5 | This module allows for generating training sets randomly for graph-based semi-supervised learning. 6 | It also allows for loading of pre-saved training sets, and to create and save training sets for future use 7 | and reproducibility of experiments. 8 | """ 9 | 10 | import numpy as np 11 | import os 12 | import sys 13 | from . import utils 14 | 15 | trainset_dir = os.path.abspath(os.path.join(os.getcwd(),'trainsets')) 16 | 17 | def load(dataset, trainset_name = ''): 18 | """Load training sets 19 | ====== 20 | 21 | Add a new dataset to graph learning by saving the data and labels. 22 | 23 | Parameters 24 | ---------- 25 | dataset : string 26 | Name of dataset. 27 | trainset_name : string (optional), default='' 28 | A modifier to uniquely identify different training sets for each dataset. 29 | """ 30 | 31 | dataFile = dataset.lower() + trainset_name.lower() +"_permutations.npz" #Change this eventually 32 | dataFile_path = os.path.join(trainset_dir, dataFile) 33 | 34 | #Check if Data directory exists 35 | if not os.path.exists(trainset_dir): 36 | os.makedirs(trainset_dir) 37 | 38 | #Download trainset if needed 39 | if not os.path.exists(dataFile_path): 40 | urlpath = 'https://github.com/jwcalder/GraphLearning/raw/master/LabelPermutations/'+dataFile 41 | utils.download_file(urlpath, dataFile_path) 42 | 43 | trainset = utils.numpy_load(dataFile_path, 'perm') 44 | 45 | return trainset 46 | 47 | def generate(labels, rate=1, num_trials=1, mask=None, dataset=None, trainset_name='', overwrite=False, seed=None): 48 | """Generate training sets 49 | ====== 50 | 51 | Generates training sets at different labeling rates over multiple trials, 52 | including features to store the training set indices to file for reproducibility. 53 | 54 | Parameters 55 | ---------- 56 | labels : numpy array, int 57 | Labels for the dataset as nonnegative integers. 58 | rate : int, float, or numpy array 59 | Controls the number of labels per class. Functionality depends on the data type. 60 | 61 | 1. A single integer is interpreted as the number of labels per class. 62 | 2. A single float in the range [0,1] is interpreted as the fraction of training data. 63 | 3. A numpy array of size (m,C) indicating different label rates, as int or float, for 64 | m different subtrials. If C=1, then the rate is extended to all classes, while if C=num classes, 65 | then the rates are interpreted on a per-class basis. 66 | num_trials : int (optional), default=1 67 | Number of training sets to generate. 68 | mask : numpy array (optional), bool, default=None 69 | If provided, then the generated training set will be selected only from points where mask=True. 70 | dataset : string (optional), default=None 71 | Name of dataset. If provided, the generated training set is saved to a file 72 | so it can be loaded later for reproducibility. 73 | trainset_name : string (optional), default='' 74 | A modifier to uniquely identify different training sets for each dataset. 75 | overwrite : bool (optional), default=False 76 | Whether to overwrite an exisiting training set file. 77 | seed : int (optional), default=None 78 | Option to seed the random number generator. 79 | 80 | Returns 81 | ------- 82 | trainset : numpy array or list of numpy arrays 83 | If m=1 and num_trials=1 then a numpy array with indices of training points is returned. 84 | Otherwise, a list of numpy arrays are returned, one for each trial. 85 | """ 86 | 87 | if seed is not None: 88 | np.random.seed(seed) 89 | 90 | unique_labels = np.unique(labels) 91 | num_per_class = np.bincount(labels) 92 | num_classes = len(unique_labels) 93 | num_points = len(labels) 94 | 95 | #Generate (m,C) integer numpy array giving number of 96 | #training points per class per trial 97 | if type(rate) == int: 98 | rate = (np.ones(num_classes)[None,:]*rate).astype(int) 99 | elif type(rate) == float: 100 | rate = (rate*num_per_class[None,:]).astype(int) 101 | elif type(rate) == np.ndarray: 102 | ratetype = rate.dtype 103 | if rate.ndim != 2: 104 | sys.exit('Must provide a 2-dimensional array for rate') 105 | if rate.shape[1] == 1: 106 | rate = rate@np.ones((1,num_classes)) 107 | if np.issubdtype(ratetype,np.integer): 108 | rate = rate.astype(int) 109 | elif np.issubdtype(ratetype,np.floating): 110 | rate = (rate*num_per_class).astype(int) 111 | else: 112 | sys.exit('Invalid numpy array type '+rate.dtype) 113 | else: 114 | sys.exit('Invalid rate type '+str(type(rate))) 115 | 116 | if mask is None: 117 | mask = np.ones(num_points,dtype=bool) 118 | 119 | #Draw training sets at random 120 | trainset = list() 121 | for k in range(num_trials): 122 | for i in range(rate.shape[0]): 123 | L = list() 124 | for j, l in enumerate(unique_labels): 125 | p = ((labels == l) & mask).astype(float) 126 | p = p/np.sum(p) 127 | L = L + np.random.choice(num_points,size=rate[i,j],p=p,replace=False).tolist() 128 | L = np.array(L) 129 | trainset.append(L) 130 | 131 | #Remove outer list if only one trial 132 | if len(trainset)==1: 133 | trainset = trainset[0] 134 | 135 | #If dataset name is provided, save permutations to file 136 | if not dataset is None: 137 | 138 | trainset = np.array(trainset,dtype=object) 139 | 140 | #data file name 141 | dataFile = dataset.lower() + trainset_name.lower() + '_permutations.npz' 142 | 143 | #Full path to file 144 | dataFile_path = os.path.join(trainset_dir, dataFile) 145 | 146 | #Check if Data directory exists 147 | if not os.path.exists(trainset_dir): 148 | os.makedirs(trainset_dir) 149 | 150 | #Save permutations to file 151 | if os.path.isfile(dataFile_path) and not overwrite: 152 | print('Training set file '+dataFile_path+' already exists. Not saving.') 153 | else: 154 | np.savez_compressed(dataFile_path,perm=trainset) 155 | 156 | return trainset 157 | 158 | 159 | -------------------------------------------------------------------------------- /images/cluster_blobs.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/images/cluster_blobs.png -------------------------------------------------------------------------------- /images/cluster_circles.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/images/cluster_circles.png -------------------------------------------------------------------------------- /images/cluster_moons.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/images/cluster_moons.png -------------------------------------------------------------------------------- /images/clustering.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/images/clustering.png -------------------------------------------------------------------------------- /kNNData/FashionMNIST_raw.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/kNNData/FashionMNIST_raw.npz -------------------------------------------------------------------------------- /kNNData/FashionMNIST_scatter.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/kNNData/FashionMNIST_scatter.npz -------------------------------------------------------------------------------- /kNNData/FashionMNIST_vae.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/kNNData/FashionMNIST_vae.npz -------------------------------------------------------------------------------- /kNNData/MNIST_raw.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/kNNData/MNIST_raw.npz -------------------------------------------------------------------------------- /kNNData/MNIST_scatter.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/kNNData/MNIST_scatter.npz -------------------------------------------------------------------------------- /kNNData/MNIST_vae.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/kNNData/MNIST_vae.npz -------------------------------------------------------------------------------- /kNNData/MNIST_vae_old.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/kNNData/MNIST_vae_old.npz -------------------------------------------------------------------------------- /kNNData/WEBKB_raw.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/kNNData/WEBKB_raw.npz -------------------------------------------------------------------------------- /kNNData/cifar10_simclr.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/kNNData/cifar10_simclr.npz -------------------------------------------------------------------------------- /kNNData/cifar10_simclr18.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/kNNData/cifar10_simclr18.npz -------------------------------------------------------------------------------- /kNNData/cifar_aet.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/kNNData/cifar_aet.npz -------------------------------------------------------------------------------- /kNNData/cifar_raw.npz: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/kNNData/cifar_raw.npz -------------------------------------------------------------------------------- /makefile: -------------------------------------------------------------------------------- 1 | all: 2 | pip install . 3 | bash generate_docs.sh 4 | -------------------------------------------------------------------------------- /pdoc/templates/_lunr_search.inc.mako: -------------------------------------------------------------------------------- 1 |
2 | 4 |
5 | 6 | 7 | 21 | 51 | -------------------------------------------------------------------------------- /pdoc/templates/config.mako: -------------------------------------------------------------------------------- 1 | <%! 2 | # Template configuration. Copy over in your template directory 3 | # (used with `--template-dir`) and adapt as necessary. 4 | # Note, defaults are loaded from this distribution file, so your 5 | # config.mako only needs to contain values you want overridden. 6 | # You can also run pdoc with `--config KEY=VALUE` to override 7 | # individual values. 8 | 9 | html_lang = 'en' 10 | show_inherited_members = False 11 | extract_module_toc_into_sidebar = True 12 | list_class_variables_in_index = True 13 | sort_identifiers = True 14 | show_type_annotations = True 15 | 16 | # Show collapsed source code block next to each item. 17 | # Disabling this can improve rendering speed of large modules. 18 | show_source_code = True 19 | 20 | # If set, format links to objects in online source code repository 21 | # according to this template. Supported keywords for interpolation 22 | # are: commit, path, start_line, end_line. 23 | #git_link_template = 'https://github.com/USER/PROJECT/blob/{commit}/{path}#L{start_line}-L{end_line}' 24 | #git_link_template = 'https://gitlab.com/USER/PROJECT/blob/{commit}/{path}#L{start_line}-L{end_line}' 25 | #git_link_template = 'https://bitbucket.org/USER/PROJECT/src/{commit}/{path}#lines-{start_line}:{end_line}' 26 | #git_link_template = 'https://CGIT_HOSTNAME/PROJECT/tree/{path}?id={commit}#n{start-line}' 27 | git_link_template = None 28 | 29 | # A prefix to use for every HTML hyperlink in the generated documentation. 30 | # No prefix results in all links being relative. 31 | link_prefix = '' 32 | 33 | # Enable syntax highlighting for code/source blocks by including Highlight.js 34 | syntax_highlighting = True 35 | 36 | # Set the style keyword such as 'atom-one-light' or 'github-gist' 37 | # Options: https://github.com/highlightjs/highlight.js/tree/master/src/styles 38 | # Demo: https://highlightjs.org/static/demo/ 39 | hljs_style = 'github' 40 | 41 | # If set, insert Google Analytics tracking code. Value is GA 42 | # tracking id (UA-XXXXXX-Y). 43 | google_analytics = '' 44 | 45 | # If set, insert Google Custom Search search bar widget above the sidebar index. 46 | # The whitespace-separated tokens represent arbitrary extra queries (at least one 47 | # must match) passed to regular Google search. Example: 48 | #google_search_query = 'inurl:github.com/USER/PROJECT site:PROJECT.github.io site:PROJECT.website' 49 | google_search_query = '' 50 | 51 | # Enable offline search using Lunr.js. For explanation of 'fuzziness' parameter, which is 52 | # added to every query word, see: https://lunrjs.com/guides/searching.html#fuzzy-matches 53 | # If 'index_docstrings' is False, a shorter index is built, indexing only 54 | # the full object reference names. 55 | #lunr_search = {'fuzziness': 1, 'index_docstrings': True} 56 | lunr_search = None 57 | 58 | # If set, render LaTeX math syntax within \(...\) (inline equations), 59 | # or within \[...\] or $$...$$ or `.. math::` (block equations) 60 | # as nicely-formatted math formulas using MathJax. 61 | # Note: in Python docstrings, either all backslashes need to be escaped (\\) 62 | # or you need to use raw r-strings. 63 | latex_math = True 64 | %> 65 | -------------------------------------------------------------------------------- /pdoc/templates/credits.mako: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/pdoc/templates/credits.mako -------------------------------------------------------------------------------- /pdoc/templates/head.mako: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/pdoc/templates/head.mako -------------------------------------------------------------------------------- /pdoc/templates/logo.mako: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/jwcalder/GraphLearning/7f18a4c8b2a46462b044df690027abb9e60efb4d/pdoc/templates/logo.mako -------------------------------------------------------------------------------- /pdoc/templates/pdf.mako: -------------------------------------------------------------------------------- 1 | <% 2 | import re 3 | import pdoc 4 | from pdoc.html_helpers import to_markdown 5 | 6 | def link(dobj: pdoc.Doc): 7 | name = dobj.qualname + ('()' if isinstance(dobj, pdoc.Function) else '') 8 | if isinstance(dobj, pdoc.External): 9 | return name 10 | return f'[{name}](#{dobj.refname} "{dobj.refname}")' 11 | 12 | def _to_md(text, module): 13 | text = to_markdown(text, docformat=docformat, module=module, link=link) 14 | # Setext H2 headings to atx H2 headings 15 | text = re.sub(r'\n(.+)\n-{3,}\n', r'\n## \1\n\n', text) 16 | # Convert admonitions into simpler paragraphs, dedent contents 17 | def paragraph_fmt(m): 18 | sub = re.sub('\n {,4}', '\n', m.group(4)) 19 | return f'{m.group(2)}**{m.group(3)}:** {sub}' 20 | text = re.sub(r'^(?P( *))!!! \w+ \"([^\"]*)\"(.*(?:\n(?P=indent) +.*)*)', 21 | paragraph_fmt, text, flags=re.MULTILINE) 22 | return text 23 | 24 | def subh(text, level=2): 25 | # Deepen heading levels so H2 becomes H4 etc. 26 | return re.sub(r'\n(#+) +(.+)\n', fr'\n{"#" * level}\1 \2\n', text) 27 | %> 28 | 29 | <%def name="title(level, string, id=None)"> 30 | <% id = f' {{#id}}' if id is not None else '' %> 31 | ${('#' * level) + ' ' + string + id} 32 | 33 | 34 | <%def name="funcdef(f)"> 35 | <% 36 | params = f.params(annotate=show_type_annotations) 37 | returns = show_type_annotations and f.return_annotation() or '' 38 | if returns: 39 | returns = ' \N{non-breaking hyphen}> ' + returns 40 | %> 41 | %if params: 42 | > ${f.funcdef()} ${f.name}( 43 | > ${',\n> '.join(params)} 44 | > )${returns} 45 | %else: 46 | > ${f.funcdef()} ${f.name}()${returns} 47 | %endif 48 | 49 | 50 | <%def name="classdef(c)"> 51 | <% params = c.params(annotate=show_type_annotations) %> 52 | %if params: 53 | > class ${c.name}( 54 | > ${',\n> '.join(params)} 55 | > ) 56 | %else: 57 | > class ${c.name} 58 | %endif 59 | 60 | 61 | <%def name="vartype(v)"> 62 | <% annot = show_type_annotations and v.type_annotation() or '' %> 63 | %if annot: 64 | Type: `${annot}` 65 | %endif 66 | 67 | 68 | --- 69 | description: | 70 | API documentation for modules: ${', '.join(m.name for m in modules)}. 71 | 72 | lang: en 73 | 74 | classoption: oneside 75 | geometry: margin=1in 76 | papersize: a4 77 | 78 | linkcolor: blue 79 | links-as-notes: true 80 | ... 81 | % for module in modules: 82 | <% 83 | submodules = module.submodules() 84 | variables = module.variables(sort=sort_identifiers) 85 | functions = module.functions(sort=sort_identifiers) 86 | classes = module.classes(sort=sort_identifiers) 87 | 88 | def to_md(text): 89 | return _to_md(text, module) 90 | %> 91 | ${title(1, ('Namespace' if module.is_namespace else 'Module') + f' `{module.name}`', module.refname)} 92 | ${module.docstring | to_md} 93 | 94 | % if submodules: 95 | ${title(2, 'Sub-modules')} 96 | % for m in submodules: 97 | * [${m.name}](#${m.refname}) 98 | % endfor 99 | % endif 100 | 101 | % if variables: 102 | ${title(2, 'Variables')} 103 | % for v in variables: 104 | ${title(3, f'Variable `{v.name}`', v.refname)} 105 | ${vartype(v)} 106 | ${v.docstring | to_md, subh, subh} 107 | % endfor 108 | % endif 109 | 110 | % if functions: 111 | ${title(2, 'Functions')} 112 | % for f in functions: 113 | ${title(3, f'Function `{f.name}`', f.refname)} 114 | 115 | ${funcdef(f)} 116 | 117 | ${f.docstring | to_md, subh, subh} 118 | % endfor 119 | % endif 120 | 121 | % if classes: 122 | ${title(2, 'Classes')} 123 | % for cls in classes: 124 | ${title(3, f'Class `{cls.name}`', cls.refname)} 125 | 126 | ${classdef(cls)} 127 | 128 | ${cls.docstring | to_md, subh} 129 | <% 130 | class_vars = cls.class_variables(show_inherited_members, sort=sort_identifiers) 131 | static_methods = cls.functions(show_inherited_members, sort=sort_identifiers) 132 | inst_vars = cls.instance_variables(show_inherited_members, sort=sort_identifiers) 133 | methods = cls.methods(show_inherited_members, sort=sort_identifiers) 134 | mro = cls.mro() 135 | subclasses = cls.subclasses() 136 | %> 137 | % if mro: 138 | ${title(4, 'Ancestors (in MRO)')} 139 | % for c in mro: 140 | * [${c.refname}](#${c.refname}) 141 | % endfor 142 | % endif 143 | 144 | % if subclasses: 145 | ${title(4, 'Descendants')} 146 | % for c in subclasses: 147 | * [${c.refname}](#${c.refname}) 148 | % endfor 149 | % endif 150 | 151 | % if class_vars: 152 | ${title(4, 'Class variables')} 153 | % for v in class_vars: 154 | ${title(5, f'Variable `{v.name}`', v.refname)} 155 | ${vartype(v)} 156 | ${v.docstring | to_md, subh, subh} 157 | % endfor 158 | % endif 159 | 160 | % if inst_vars: 161 | ${title(4, 'Instance variables')} 162 | % for v in inst_vars: 163 | ${title(5, f'Variable `{v.name}`', v.refname)} 164 | ${vartype(v)} 165 | ${v.docstring | to_md, subh, subh} 166 | % endfor 167 | % endif 168 | 169 | % if static_methods: 170 | ${title(4, 'Static methods')} 171 | % for f in static_methods: 172 | ${title(5, f'`Method {f.name}`', f.refname)} 173 | 174 | ${funcdef(f)} 175 | 176 | ${f.docstring | to_md, subh, subh} 177 | % endfor 178 | % endif 179 | 180 | % if methods: 181 | ${title(4, 'Methods')} 182 | % for f in methods: 183 | ${title(5, f'Method `{f.name}`', f.refname)} 184 | 185 | ${funcdef(f)} 186 | 187 | ${f.docstring | to_md, subh, subh} 188 | % endfor 189 | % endif 190 | % endfor 191 | % endif 192 | 193 | ##\## for module in modules: 194 | % endfor 195 | 196 | ----- 197 | Generated by *pdoc* ${pdoc.__version__} (). 198 | -------------------------------------------------------------------------------- /pdoc/templates/search.mako: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | 6 | Search 7 | 8 | 9 | 28 | 29 | 30 | 31 | 34 |
35 |

36 |
    37 |
    38 |
    39 |

    Search results provided by Lunr.js

    40 |
    41 | 42 | 43 | 44 | 142 | 143 | -------------------------------------------------------------------------------- /pdoc/templates/text.mako: -------------------------------------------------------------------------------- 1 | ## Define mini-templates for each portion of the doco. 2 | 3 | <%! 4 | def indent(s, spaces=4): 5 | new = s.replace('\n', '\n' + ' ' * spaces) 6 | return ' ' * spaces + new.strip() 7 | %> 8 | 9 | <%def name="deflist(s)">:${indent(s)[1:]} 10 | 11 | <%def name="h3(s)">### ${s} 12 | 13 | 14 | <%def name="function(func)" buffered="True"> 15 | <% 16 | returns = show_type_annotations and func.return_annotation() or '' 17 | if returns: 18 | returns = ' \N{non-breaking hyphen}> ' + returns 19 | %> 20 | `${func.name}(${", ".join(func.params(annotate=show_type_annotations))})${returns}` 21 | ${func.docstring | deflist} 22 | 23 | 24 | <%def name="variable(var)" buffered="True"> 25 | <% 26 | annot = show_type_annotations and var.type_annotation() or '' 27 | if annot: 28 | annot = ': ' + annot 29 | %> 30 | `${var.name}${annot}` 31 | ${var.docstring | deflist} 32 | 33 | 34 | <%def name="class_(cls)" buffered="True"> 35 | `${cls.name}(${", ".join(cls.params(annotate=show_type_annotations))})` 36 | ${cls.docstring | deflist} 37 | <% 38 | class_vars = cls.class_variables(show_inherited_members, sort=sort_identifiers) 39 | static_methods = cls.functions(show_inherited_members, sort=sort_identifiers) 40 | inst_vars = cls.instance_variables(show_inherited_members, sort=sort_identifiers) 41 | methods = cls.methods(show_inherited_members, sort=sort_identifiers) 42 | mro = cls.mro() 43 | subclasses = cls.subclasses() 44 | %> 45 | % if mro: 46 | ${h3('Ancestors (in MRO)')} 47 | % for c in mro: 48 | * ${c.refname} 49 | % endfor 50 | 51 | % endif 52 | % if subclasses: 53 | ${h3('Descendants')} 54 | % for c in subclasses: 55 | * ${c.refname} 56 | % endfor 57 | 58 | % endif 59 | % if class_vars: 60 | ${h3('Class variables')} 61 | % for v in class_vars: 62 | ${variable(v) | indent} 63 | 64 | % endfor 65 | % endif 66 | % if static_methods: 67 | ${h3('Static methods')} 68 | % for f in static_methods: 69 | ${function(f) | indent} 70 | 71 | % endfor 72 | % endif 73 | % if inst_vars: 74 | ${h3('Instance variables')} 75 | % for v in inst_vars: 76 | ${variable(v) | indent} 77 | 78 | % endfor 79 | % endif 80 | % if methods: 81 | ${h3('Methods')} 82 | % for m in methods: 83 | ${function(m) | indent} 84 | 85 | % endfor 86 | % endif 87 | 88 | 89 | ## Start the output logic for an entire module. 90 | 91 | <% 92 | variables = module.variables(sort=sort_identifiers) 93 | classes = module.classes(sort=sort_identifiers) 94 | functions = module.functions(sort=sort_identifiers) 95 | submodules = module.submodules() 96 | heading = 'Namespace' if module.is_namespace else 'Module' 97 | %> 98 | 99 | ${heading} ${module.name} 100 | =${'=' * (len(module.name) + len(heading))} 101 | ${module.docstring} 102 | 103 | 104 | % if submodules: 105 | Sub-modules 106 | ----------- 107 | % for m in submodules: 108 | * ${m.name} 109 | % endfor 110 | % endif 111 | 112 | % if variables: 113 | Variables 114 | --------- 115 | % for v in variables: 116 | ${variable(v)} 117 | 118 | % endfor 119 | % endif 120 | 121 | % if functions: 122 | Functions 123 | --------- 124 | % for f in functions: 125 | ${function(f)} 126 | 127 | % endfor 128 | % endif 129 | 130 | % if classes: 131 | Classes 132 | ------- 133 | % for c in classes: 134 | ${class_(c)} 135 | 136 | % endfor 137 | % endif 138 | -------------------------------------------------------------------------------- /pyproject.toml: -------------------------------------------------------------------------------- 1 | [build-system] 2 | requires = ["setuptools","numpy"] 3 | build-backend = "setuptools.build_meta" 4 | 5 | 6 | [tool.setuptools] 7 | packages = ['graphlearning'] 8 | 9 | 10 | [project] 11 | name = "graphlearning" 12 | version = "1.7.2" 13 | authors = [ 14 | { name="Jeff Calder", email="jwcalder@umn.edu" }, 15 | ] 16 | description = "Python package for graph-based clustering and semi-supervised learning" 17 | readme = "README.md" 18 | license = {text = "MIT"} 19 | requires-python = ">=3.6" 20 | classifiers = [ 21 | "Programming Language :: Python :: 3", 22 | "License :: OSI Approved :: MIT License", 23 | "Operating System :: OS Independent", 24 | ] 25 | dependencies = [ 26 | 'numpy', 27 | 'scipy', 28 | 'scikit-learn', 29 | 'matplotlib' 30 | ] 31 | 32 | [project.urls] 33 | "Homepage" = "https://github.com/jwcalder/GraphLearning" 34 | "Bug Tracker" = "https://github.com/jwcalder/GraphLearning/issues" 35 | -------------------------------------------------------------------------------- /requirements.txt: -------------------------------------------------------------------------------- 1 | numpy 2 | scipy 3 | matplotlib 4 | scikit-learn 5 | -------------------------------------------------------------------------------- /setup.py: -------------------------------------------------------------------------------- 1 | from setuptools import setup, Extension 2 | import numpy 3 | 4 | setup_args = dict( 5 | ext_modules=[Extension('graphlearning.cextensions', 6 | sources=['c_code/cextensions.cpp', 7 | 'c_code/lp_iterate.cpp', 8 | 'c_code/hjsolvers.cpp', 9 | 'c_code/memory_allocation.cpp', 10 | 'c_code/mnist_benchmark.cpp', 11 | 'c_code/mbo_convolution.cpp', 12 | 'c_code/tsne.cpp', 13 | 'c_code/sptree.cpp'], 14 | include_dirs=[numpy.get_include()], 15 | #extra_compile_args = ['-Ofast','-std=c11'], 16 | extra_compile_args = ['-Ofast'], 17 | extra_link_args = ['-lm'])]) 18 | 19 | setup(**setup_args) 20 | --------------------------------------------------------------------------------