├── figures └── exp.png ├── models ├── motorLSTM.pth.tar └── p300LSTM.pth.tar ├── .gitignore ├── README.md └── 03 - Asymmetry + DEAP.ipynb /figures/exp.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/chaklam-silpasuwanchai/Python-for-EEG/HEAD/figures/exp.png -------------------------------------------------------------------------------- /models/motorLSTM.pth.tar: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/chaklam-silpasuwanchai/Python-for-EEG/HEAD/models/motorLSTM.pth.tar -------------------------------------------------------------------------------- /models/p300LSTM.pth.tar: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/chaklam-silpasuwanchai/Python-for-EEG/HEAD/models/p300LSTM.pth.tar -------------------------------------------------------------------------------- /.gitignore: -------------------------------------------------------------------------------- 1 | **/*.ipynb_checkpoints/ 2 | # Byte-compiled / optimized / DLL files 3 | __pycache__/ 4 | *.py[cod] 5 | 6 | # C extensions 7 | *.so 8 | 9 | # Distribution / packaging 10 | bin/ 11 | build/ 12 | develop-eggs/ 13 | dist/ 14 | eggs/ 15 | lib/ 16 | lib64/ 17 | parts/ 18 | sdist/ 19 | var/ 20 | *.egg-info/ 21 | .installed.cfg 22 | *.egg 23 | 24 | # Installer logs 25 | pip-log.txt 26 | pip-delete-this-directory.txt 27 | 28 | # Unit test / coverage reports 29 | .tox/ 30 | .coverage 31 | .cache 32 | nosetests.xml 33 | coverage.xml 34 | 35 | # Translations 36 | *.mo 37 | 38 | # Mr Developer 39 | .mr.developer.cfg 40 | .project 41 | .pydevproject 42 | 43 | # Rope 44 | .ropeproject 45 | 46 | # Django stuff: 47 | *.log 48 | *.pot 49 | 50 | # Sphinx documentation 51 | docs/_build/ 52 | 53 | # VScode 54 | /.vscode 55 | # Jupytor 56 | /.ipynb_checkpoints 57 | 58 | AkraradetSolution 59 | 60 | .DS_Store 61 | ._.DS_Store 62 | **/.DS_Store 63 | **/._.DS_Store 64 | 65 | Quiz/ 66 | 67 | .data 68 | ._.data 69 | **/.data 70 | **/._.data 71 | 72 | data 73 | ._.data 74 | **/data 75 | **/data 76 | 77 | models 78 | ._.models 79 | **/models 80 | **/models 81 | 82 | # Pipenv 83 | .venv 84 | Pipfile.lock 85 | 86 | # 05 - Signal Processing for EEG 87 | Lectures/05 - Signal Processing for EEG/dataset/* 88 | Lectures/me 89 | 90 | cache -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # Python-For-EEG 2 | 3 | Here is a simple repository in which I demonstrate basic analysis on EEG signal. Topics covered: 4 | 5 | Credits: I got help from my Ph.D. student - Mr. Akraradet (https://github.com/akraradets) on asymmetry and my Master student - Mr. Nutapol (https://github.com/nutapol97) on motor imagery. 6 | 7 | ## Topics 8 | 9 | **Time-Based** [Datasets: P300 signal] 10 | 11 | 1. Event-related potentials and 1-dimensional convolution 12 | 2. Long short-term memory 13 | 14 | **Frequency-Based** [Datasets: DEAP and SSVEP] 15 | 16 | 3. Spectral analysis and alpha asymmetry 17 | 4. Canonical correlation analysis 18 | 19 | **Component-Based** [Datasets: Motor Imagery] 20 | 21 | 1. Event-related desynchronization 22 | 23 | 24 | ## Datasets 25 | 26 | All datasets can be downloaded at: [Google Drive](https://drive.google.com/drive/folders/1q_UbAIP1yPCkIjYCMIaJWG2cBn0K4nfa?usp=sharing) (except the DEAP dataset) 27 | 28 | 1. P300: This is my personal P300 signal I acquired through my OpenBCI headset while visually attending to 36 alphabetical targets. When the flash flickers at the target I am looking at, there will be a signal rise at 300ms after the stimulus onset. In the dataset, the columns are timestamps (128 samples per second), 16 channels of 'Fp1', 'Fp2', 'F7', 'F3', 'F4', 'F8', 'C3', 'Cz', 'C4', 'T5', 'P3', 'P4', 'T6', 'POz', 'O1', 'O2', and the marker indicating the events. The marker format will be explained more in the notebook. 29 | 30 | 2. DEAP dataset: It's basically a dataset regarding participants watching some 1-min emotional videos while wearing 32 EEG channels. For more details, read https://www.eecs.qmul.ac.uk/mmv/datasets/deap/ 31 | 32 | 3. SSVEP: Here we record users looking at three different circles flickering at 6, 10, and 15Hz respectively. We will be classifying the signals using filterbank canonical correlation analysis. 33 | 34 | 4. Motor Imagery: Here we record one user performing imagined left and right movements. We shall explore event-related desynchronization for decoding the classes. 35 | -------------------------------------------------------------------------------- /03 - Asymmetry + DEAP.ipynb: -------------------------------------------------------------------------------- 1 | { 2 | "cells": [ 3 | { 4 | "cell_type": "markdown", 5 | "id": "1b871aa0", 6 | "metadata": {}, 7 | "source": [ 8 | "# Asymmetry + DEAP (emotion)\n", 9 | "\n", 10 | "In this tutorial, we will be working on the DEAP dataset, a benchmark EEG emotion recognition dataset. \n", 11 | "\n", 12 | "In this dataset, there is a total of 32 participants, where each participant watches 40 1-minute videos. Thus s01.dat is holding 40 batches. For simplicity, we shall only use the first participant, i.e., s01.dat. \n", 13 | "\n", 14 | "You can download the dataset by googling and ask permission from the owner by filling a form.\n", 15 | "\n", 16 | "Looking in each dat file (e.g., s01), it contains the data and label\n", 17 | "- Data ----- 40 x 40 x 8064 [\tvideo/batches x channel x samples ]\n", 18 | "- Label ---- 40 x 4 \n", 19 | "\n", 20 | "Out of 40 channels, 32 channels were of EEG, and the rest of 8 of them from other sensors such as EOG (see the section 6.1 of the original paper). We shall only extract the first 32 channels. For the 8064, since the data is downsampled to 128Hz, thus one second contains around 128 samples, thus in one minute which is 60 seconds, it will be roughly 7680 samples. The paper did not really talk a lot but it is likely there is another 1.5 seconds before and after which total to 8064 samples (128 Hz * 63 seconds).\n", 21 | "\n", 22 | "The four labels correspond to valence, arousal, liking, and dominance, in this order. We will only use valence and arousal, thus index 0 and 1 of the labels will be extracted.\n", 23 | "\n", 24 | "For more information, see https://www.eecs.qmul.ac.uk/mmv/datasets/deap/readme.html" 25 | ] 26 | }, 27 | { 28 | "cell_type": "code", 29 | "execution_count": 1, 30 | "id": "49033f3c", 31 | "metadata": {}, 32 | "outputs": [], 33 | "source": [ 34 | "import pickle\n", 35 | "import os\n", 36 | "\n", 37 | "import mne\n", 38 | "from mne import create_info\n", 39 | "from mne.io import RawArray\n", 40 | "\n", 41 | "import numpy as np\n", 42 | "import pandas as pd\n", 43 | "import seaborn as sns\n", 44 | "import matplotlib.pyplot as plt\n", 45 | "\n", 46 | "import torch\n", 47 | "import torch.nn as nn" 48 | ] 49 | }, 50 | { 51 | "cell_type": "code", 52 | "execution_count": 2, 53 | "id": "e55a7ab1", 54 | "metadata": {}, 55 | "outputs": [], 56 | "source": [ 57 | "np.set_printoptions(precision=4)" 58 | ] 59 | }, 60 | { 61 | "cell_type": "code", 62 | "execution_count": 3, 63 | "id": "4d447abd", 64 | "metadata": {}, 65 | "outputs": [ 66 | { 67 | "name": "stdout", 68 | "output_type": "stream", 69 | "text": [ 70 | "cpu\n" 71 | ] 72 | } 73 | ], 74 | "source": [ 75 | "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n", 76 | "print(device)" 77 | ] 78 | }, 79 | { 80 | "cell_type": "code", 81 | "execution_count": 4, 82 | "id": "e9f93d9f", 83 | "metadata": {}, 84 | "outputs": [], 85 | "source": [ 86 | "# torch.cuda.get_device_name(0)" 87 | ] 88 | }, 89 | { 90 | "cell_type": "markdown", 91 | "id": "7042a8e2", 92 | "metadata": {}, 93 | "source": [ 94 | "## 1. Loading the dataset\n", 95 | "\n", 96 | "Let's first create a simple dataset loader. The code is explained using comments and is quite self-explanatory." 97 | ] 98 | }, 99 | { 100 | "cell_type": "code", 101 | "execution_count": 5, 102 | "id": "bf31734b", 103 | "metadata": {}, 104 | "outputs": [], 105 | "source": [ 106 | "class Dataset(torch.utils.data.Dataset):\n", 107 | " \n", 108 | " def __init__(self, path, stim):\n", 109 | " _, _, filenames = next(os.walk(path))\n", 110 | " filenames = sorted(filenames)\n", 111 | " all_data = []\n", 112 | " all_label = []\n", 113 | " for dat in filenames:\n", 114 | " temp = pickle.load(open(os.path.join(path,dat), 'rb'), encoding='latin1')\n", 115 | " all_data.append(temp['data'])\n", 116 | " \n", 117 | " if stim == \"Valence\":\n", 118 | " all_label.append(temp['labels'][:,:1]) #the first index is valence\n", 119 | " elif stim == \"Arousal\":\n", 120 | " all_label.append(temp['labels'][:,1:2]) # Arousal #the second index is arousal\n", 121 | " \n", 122 | " self.data = np.vstack(all_data)[:, :32, ] #shape: (40, 32, 8064) ==> (samples, channels, samples)\n", 123 | " self.label = np.vstack(all_label) #(40, ) ==> 1280 samples, each with a unique label (depend on the param \"stim\")\n", 124 | " \n", 125 | " del temp, all_data, all_label\n", 126 | "\n", 127 | " def __len__(self):\n", 128 | " return self.data.shape[0]\n", 129 | "\n", 130 | " def __getitem__(self, idx):\n", 131 | " single_data = self.data[idx]\n", 132 | " single_label = (self.label[idx] > 5).astype(float) #convert the scale to either 0 or 1 (to classification problem)\n", 133 | " \n", 134 | " batch = {\n", 135 | " 'data': torch.Tensor(single_data),\n", 136 | " 'label': torch.Tensor(single_label)\n", 137 | " }\n", 138 | " \n", 139 | " return batch" 140 | ] 141 | }, 142 | { 143 | "cell_type": "markdown", 144 | "id": "e1a8800b", 145 | "metadata": {}, 146 | "source": [ 147 | "Let's try load the dataset." 148 | ] 149 | }, 150 | { 151 | "cell_type": "code", 152 | "execution_count": 6, 153 | "id": "06593774", 154 | "metadata": {}, 155 | "outputs": [], 156 | "source": [ 157 | "path = \"data/deap\" #create a folder \"data\", and inside put s01.dat,....,s32.dat inside from the preprocessed folder from the DEAP dataset" 158 | ] 159 | }, 160 | { 161 | "cell_type": "code", 162 | "execution_count": 7, 163 | "id": "141357b6", 164 | "metadata": {}, 165 | "outputs": [], 166 | "source": [ 167 | "dataset_valence = Dataset(path, \"Valence\")\n", 168 | "dataset_arousal = Dataset(path, \"Arousal\")" 169 | ] 170 | }, 171 | { 172 | "cell_type": "markdown", 173 | "id": "22b3d4f2", 174 | "metadata": {}, 175 | "source": [ 176 | "We can try look at one sample using the index. This is automatically mapped to the __getitem__ function in the Dataset class." 177 | ] 178 | }, 179 | { 180 | "cell_type": "code", 181 | "execution_count": 8, 182 | "id": "94f0f77d", 183 | "metadata": {}, 184 | "outputs": [ 185 | { 186 | "data": { 187 | "text/plain": [ 188 | "{'data': tensor([[ 0.9482, 1.6533, 3.0137, ..., -2.8265, -4.4772, -3.6769],\n", 189 | " [ 0.1247, 1.3901, 1.8351, ..., -2.9870, -6.2878, -4.4743],\n", 190 | " [-2.2165, 2.2920, 2.7464, ..., -2.6371, -7.4065, -6.7559],\n", 191 | " ...,\n", 192 | " [-3.5572, -1.2603, -3.0996, ..., 6.6889, 6.9123, 5.6690],\n", 193 | " [-1.2373, -0.9414, 1.0280, ..., 2.6402, 6.5746, 7.2085],\n", 194 | " [ 0.3723, 2.0762, 4.4652, ..., 2.2396, 3.1826, 4.7081]]),\n", 195 | " 'label': tensor([1.])}" 196 | ] 197 | }, 198 | "execution_count": 8, 199 | "metadata": {}, 200 | "output_type": "execute_result" 201 | } 202 | ], 203 | "source": [ 204 | "dataset_valence[0]" 205 | ] 206 | }, 207 | { 208 | "cell_type": "code", 209 | "execution_count": 9, 210 | "id": "b339d328", 211 | "metadata": {}, 212 | "outputs": [ 213 | { 214 | "name": "stdout", 215 | "output_type": "stream", 216 | "text": [ 217 | "Shape of data: torch.Size([32, 8064])\n", 218 | "Shape of label: torch.Size([1])\n" 219 | ] 220 | } 221 | ], 222 | "source": [ 223 | "print(\"Shape of data: \", dataset_valence[0]['data'].shape) #32 channels of data, 8064 samples in 1 minute\n", 224 | "print(\"Shape of label: \", dataset_valence[0]['label'].shape) #just 1 single label; 0 or 1" 225 | ] 226 | }, 227 | { 228 | "cell_type": "markdown", 229 | "id": "e53665b7", 230 | "metadata": {}, 231 | "source": [ 232 | "Let's try to look at our data and label distribution." 233 | ] 234 | }, 235 | { 236 | "cell_type": "code", 237 | "execution_count": 10, 238 | "id": "c82238f4", 239 | "metadata": {}, 240 | "outputs": [], 241 | "source": [ 242 | "data = dataset_valence[:]['data']\n", 243 | "label = dataset_valence[:]['label']" 244 | ] 245 | }, 246 | { 247 | "cell_type": "code", 248 | "execution_count": 11, 249 | "id": "1f56c42e", 250 | "metadata": {}, 251 | "outputs": [ 252 | { 253 | "data": { 254 | "text/plain": [ 255 | "torch.Size([40, 32, 8064])" 256 | ] 257 | }, 258 | "execution_count": 11, 259 | "metadata": {}, 260 | "output_type": "execute_result" 261 | } 262 | ], 263 | "source": [ 264 | "#so we got 40 trials (40 videos, each with 32 channels of data, each video contains 8064 EEG samples)\n", 265 | "data.shape " 266 | ] 267 | }, 268 | { 269 | "cell_type": "code", 270 | "execution_count": 12, 271 | "id": "0c5c274d", 272 | "metadata": {}, 273 | "outputs": [ 274 | { 275 | "data": { 276 | "text/plain": [ 277 | "torch.Size([40, 1])" 278 | ] 279 | }, 280 | "execution_count": 12, 281 | "metadata": {}, 282 | "output_type": "execute_result" 283 | } 284 | ], 285 | "source": [ 286 | "#so we got 40 labels, i.e., one label per video\n", 287 | "label.shape " 288 | ] 289 | }, 290 | { 291 | "cell_type": "markdown", 292 | "id": "9cd55e43", 293 | "metadata": {}, 294 | "source": [ 295 | "Let's count how many 0 and 1 in the valence dataset, to see if there is some imbalance." 296 | ] 297 | }, 298 | { 299 | "cell_type": "code", 300 | "execution_count": 13, 301 | "id": "a57f6656", 302 | "metadata": {}, 303 | "outputs": [ 304 | { 305 | "name": "stdout", 306 | "output_type": "stream", 307 | "text": [ 308 | "Labels 1 in valence dataset: 19\n", 309 | "Labels 0 in valence dataset: 21\n" 310 | ] 311 | } 312 | ], 313 | "source": [ 314 | "cond_1 = label == 1\n", 315 | "cond_0 = label == 0\n", 316 | "\n", 317 | "print(\"Labels 1 in valence dataset: \", len(label[cond_1]))\n", 318 | "print(\"Labels 0 in valence dataset: \", len(label[cond_0]))" 319 | ] 320 | }, 321 | { 322 | "cell_type": "markdown", 323 | "id": "d9b5925d", 324 | "metadata": {}, 325 | "source": [ 326 | "Let's also count in the arousal dataset, to see if there is some imbalance." 327 | ] 328 | }, 329 | { 330 | "cell_type": "code", 331 | "execution_count": 14, 332 | "id": "52c09d42", 333 | "metadata": {}, 334 | "outputs": [ 335 | { 336 | "name": "stdout", 337 | "output_type": "stream", 338 | "text": [ 339 | "Labels 1 in arousal dataset: 19\n", 340 | "Labels 0 in arousal dataset: 21\n" 341 | ] 342 | } 343 | ], 344 | "source": [ 345 | "cond_1 = label == 1\n", 346 | "cond_0 = label == 0\n", 347 | "\n", 348 | "print(\"Labels 1 in arousal dataset: \", len(label[cond_1]))\n", 349 | "print(\"Labels 0 in arousal dataset: \", len(label[cond_0]))" 350 | ] 351 | }, 352 | { 353 | "cell_type": "markdown", 354 | "id": "5c97dcdf", 355 | "metadata": {}, 356 | "source": [ 357 | "## 2. Artifact Removal\n", 358 | "\n", 359 | "Note that since the data is already preprocessed by the authors, we don't have to do anything more, but it's very natural for us to do preprocessing, e.g., min-max normalization, notch filters, band pass filters, etc.\n", 360 | "\n", 361 | "Base on the information given in the DEAP website, the preprocessed data has done the following;\n", 362 | "\n", 363 | "1. The data was downsampled to 128Hz.\n", 364 | "2. EOG artefacts were removed as in [1].\n", 365 | "3. A bandpass frequency filter from 4.0-45.0Hz was applied.\n", 366 | "4. The data was averaged to the common reference.\n", 367 | "5. The EEG channels were reordered so that they all follow the Geneva order as above.\n", 368 | "6. The data was segmented into 60 second trials and a 3 second pre-trial baseline removed.\n", 369 | "7. The trials were reordered from presentation order to video (Experiment_id) order." 370 | ] 371 | }, 372 | { 373 | "cell_type": "markdown", 374 | "id": "4b7a4027", 375 | "metadata": {}, 376 | "source": [ 377 | "## 3. Epoching" 378 | ] 379 | }, 380 | { 381 | "cell_type": "markdown", 382 | "id": "fbc28fa8", 383 | "metadata": {}, 384 | "source": [ 385 | "Given that DEAP dataset already provides their data in epochs, we can simply put these epochs as mne epochs object by using `mne.EpochsArray`" 386 | ] 387 | }, 388 | { 389 | "cell_type": "code", 390 | "execution_count": 15, 391 | "id": "4fc1eb36", 392 | "metadata": {}, 393 | "outputs": [ 394 | { 395 | "name": "stdout", 396 | "output_type": "stream", 397 | "text": [ 398 | "Not setting metadata\n", 399 | "40 matching events found\n", 400 | "No baseline correction applied\n", 401 | "0 projection items activated\n", 402 | "----\n", 403 | "epochs._data.shape=(40, 32, 8064)\n", 404 | "----\n" 405 | ] 406 | }, 407 | { 408 | "data": { 409 | "text/html": [ 410 | "\n", 411 | " \n", 412 | " \n", 413 | " \n", 414 | " \n", 415 | " \n", 416 | " \n", 417 | " \n", 418 | " \n", 419 | " \n", 420 | " \n", 421 | " \n", 422 | " \n", 423 | " \n", 424 | " \n", 425 | " \n", 426 | " \n", 427 | " \n", 428 | " \n", 429 | "
Number of events40
Events1: 40
Time range0.000 – 62.992 sec
Baselineoff
" 430 | ], 431 | "text/plain": [ 432 | "" 434 | ] 435 | }, 436 | "execution_count": 15, 437 | "metadata": {}, 438 | "output_type": "execute_result" 439 | } 440 | ], 441 | "source": [ 442 | "#meta data\n", 443 | "ch_names = ['Fp1','AF3','F3','F7','FC5','FC1','C3','T7','CP5','CP1','P3','P7','PO3','O1','Oz','Pz','Fp2','AF4','Fz','F4','F8','FC6','FC2','Cz','C4','T8','CP6','CP2','P4','P8','PO4','O2']\n", 444 | "ch_types = ['eeg'] * len(ch_names)\n", 445 | "sfreq = 128 #Hz\n", 446 | "info = mne.create_info(ch_names=ch_names, ch_types=ch_types, sfreq=sfreq)\n", 447 | "\n", 448 | "#create epochs\n", 449 | "epochs = mne.EpochsArray(dataset_valence[:]['data'],info)\n", 450 | "\n", 451 | "#set headset mapping\n", 452 | "epochs.set_montage('standard_1020')\n", 453 | "print('----')\n", 454 | "# You can access the original data using <>._data \n", 455 | "print(f\"{epochs._data.shape=}\")\n", 456 | "print('----')\n", 457 | "epochs" 458 | ] 459 | }, 460 | { 461 | "cell_type": "markdown", 462 | "id": "f306ba5a", 463 | "metadata": {}, 464 | "source": [ 465 | "Here is some short summary regarding mne data structures:\n", 466 | "\n", 467 | "https://mne.tools/stable/most_used_classes.html\n", 468 | "\n", 469 | "1. `MNE.io.Raw`\n", 470 | " - The `MNE.io.Raw` is you entire session. In this class, you can perform a preprocess for the entire session like `filter`\n", 471 | "2. `MNE.Epochs`\n", 472 | " - The Epochs is an idea for segment the entire session to smaller windows. You can epoching you session in any how you like. The most common epoching is to follow the `marker/event` channel\n", 473 | "3. `MNE.Evoked`\n", 474 | " - Evoked is a data that you obtain from each epoch. You can think of the Epochs is an Array of Evoked.\n", 475 | "4. `MNE.Spectrum`\n", 476 | " - for all three objects above, once you called `compute_psd()`, it became a MNE Spectrum instance, in which you can retrieve power of different frequencies." 477 | ] 478 | }, 479 | { 480 | "cell_type": "markdown", 481 | "id": "1a3facae", 482 | "metadata": {}, 483 | "source": [ 484 | "## 4. EDA - Spectral Analysis\n", 485 | "\n", 486 | "Here we focus on feature engineering using spectral analysis. Spectral analysis here refers to the analysis of theta (4 - 8 Hz), alpha (8 - 12 Hz), beta (12 - 30 Hz), and gamma (30 - 64 Hz). \n", 487 | "\n", 488 | "Spectral analysis is a very basic and must-do analysis for emotions/cognitions/resting state since it is a common knowledge with abundant evidence that our emotion/cognition change how our brain signals oscillate. For example, when we are calm, alpha is relatively high, likewise, when we are attentive, beta is relatively high and alpha becomes relatively lower.\n", 489 | "\n", 490 | "MNE has functions to calulate PSD. There are multiple methods to get PSD from a signal (and other Voltage-to-Frequency domain convert technique).\n", 491 | "\n", 492 | "A welch method is widely used.\n" 493 | ] 494 | }, 495 | { 496 | "cell_type": "code", 497 | "execution_count": 16, 498 | "id": "ef6628cf", 499 | "metadata": {}, 500 | "outputs": [ 501 | { 502 | "name": "stdout", 503 | "output_type": "stream", 504 | "text": [ 505 | "Effective window size : 1.000 (s)\n", 506 | "(1, 32, 65) (65,)\n" 507 | ] 508 | } 509 | ], 510 | "source": [ 511 | "epoch = epochs.__getitem__(0) #first epoch\n", 512 | "spectrum = epoch.compute_psd(method=\"welch\", n_fft=128) #get Spectrum object from epochs\n", 513 | "psds, freqs = spectrum.get_data(return_freqs=True)\n", 514 | "print(psds.shape, freqs.shape)" 515 | ] 516 | }, 517 | { 518 | "cell_type": "markdown", 519 | "id": "de529ea1", 520 | "metadata": {}, 521 | "source": [ 522 | "`psd_welch` returns you two objects, PSDs and FREQs.\n", 523 | "\n", 524 | "The shape of PSDs is (n_epochs,n_channels,n_freqs) and FREQs is (n_freqs)\n", 525 | "\n", 526 | "As you can see, the range of frequency returned from the function is [0-64] because of nyquist theory.\n", 527 | "\n", 528 | "The basic is you can only obtain frequncy range up to sampling_rate/2. In this data, the sampling rate/freq is 128Hz, therefor the maximun frequncy is 128/2 = 64 Hz.\n", 529 | "\n", 530 | "This is specified using the `n_fft` parameter as 128.\n", 531 | "\n", 532 | "Now let's retrieve the power for all 40 epochs, by putting epochs instead of epoch. In addition, we shall apply a log to the psds, which is a common way to enhance the visibility of the spectrum." 533 | ] 534 | }, 535 | { 536 | "cell_type": "code", 537 | "execution_count": 17, 538 | "id": "256d085d", 539 | "metadata": {}, 540 | "outputs": [ 541 | { 542 | "name": "stdout", 543 | "output_type": "stream", 544 | "text": [ 545 | "Effective window size : 1.000 (s)\n", 546 | "(40, 32, 65) (65,)\n" 547 | ] 548 | } 549 | ], 550 | "source": [ 551 | "spectrum = epochs.compute_psd(method=\"welch\", n_fft=128, ) #get Spectrum object from epochs\n", 552 | "psds, freqs = spectrum.get_data(return_freqs=True)\n", 553 | "psds = 10 * np.log10(psds)\n", 554 | "print(psds.shape, freqs.shape)" 555 | ] 556 | }, 557 | { 558 | "cell_type": "markdown", 559 | "id": "e7c7f057", 560 | "metadata": {}, 561 | "source": [ 562 | "Let's try to plot different bands, and see their difference" 563 | ] 564 | }, 565 | { 566 | "cell_type": "code", 567 | "execution_count": 18, 568 | "id": "a6f2fad3", 569 | "metadata": {}, 570 | "outputs": [ 571 | { 572 | "data": { 573 | "text/plain": [ 574 | "(40, 65)" 575 | ] 576 | }, 577 | "execution_count": 18, 578 | "metadata": {}, 579 | "output_type": "execute_result" 580 | } 581 | ], 582 | "source": [ 583 | "#1. first we take the mean of all channels\n", 584 | "psds = psds.mean(axis=1)\n", 585 | "psds.shape #(40, 65)" 586 | ] 587 | }, 588 | { 589 | "cell_type": "markdown", 590 | "id": "a8c91f09", 591 | "metadata": {}, 592 | "source": [ 593 | "Next we gonna the `(_,65)` to `(_,5)`, i.e., delta (0-4), theta (4 - 8 Hz), alpha (8 - 12 Hz), beta (12 - 30 Hz), and gamma (30 - 64 Hz)\n", 594 | "\n", 595 | "**Note that the range of delta,theta,... is different from paper to paper*" 596 | ] 597 | }, 598 | { 599 | "cell_type": "code", 600 | "execution_count": 19, 601 | "id": "f8be7d97", 602 | "metadata": {}, 603 | "outputs": [ 604 | { 605 | "data": { 606 | "text/plain": [ 607 | "((40, 5), (40, 5), (40, 5), (40, 19), (40, 35))" 608 | ] 609 | }, 610 | "execution_count": 19, 611 | "metadata": {}, 612 | "output_type": "execute_result" 613 | } 614 | ], 615 | "source": [ 616 | "delta = psds[:, :5]\n", 617 | "theta = psds[:, 4:9]\n", 618 | "alpha = psds[:, 8:13]\n", 619 | "beta = psds[:, 12:31]\n", 620 | "gamma = psds[:, 30:]\n", 621 | "\n", 622 | "delta.shape, theta.shape, alpha.shape, beta.shape, gamma.shape" 623 | ] 624 | }, 625 | { 626 | "cell_type": "markdown", 627 | "id": "a62575a9", 628 | "metadata": {}, 629 | "source": [ 630 | "Then we can just mean all the power " 631 | ] 632 | }, 633 | { 634 | "cell_type": "code", 635 | "execution_count": 20, 636 | "id": "9734dc67", 637 | "metadata": {}, 638 | "outputs": [ 639 | { 640 | "data": { 641 | "text/plain": [ 642 | "((40, 1), (40, 1), (40, 1), (40, 1), (40, 1))" 643 | ] 644 | }, 645 | "execution_count": 20, 646 | "metadata": {}, 647 | "output_type": "execute_result" 648 | } 649 | ], 650 | "source": [ 651 | "delta = delta.mean(axis=1).reshape(-1, 1)\n", 652 | "theta = theta.mean(axis=1).reshape(-1, 1)\n", 653 | "alpha = alpha.mean(axis=1).reshape(-1, 1)\n", 654 | "beta = beta.mean(axis=1).reshape(-1, 1)\n", 655 | "gamma = gamma.mean(axis=1).reshape(-1, 1)\n", 656 | "\n", 657 | "delta.shape, theta.shape, alpha.shape, beta.shape, gamma.shape" 658 | ] 659 | }, 660 | { 661 | "cell_type": "markdown", 662 | "id": "9d147715", 663 | "metadata": {}, 664 | "source": [ 665 | "For plotting the bar graph, let's stack them together" 666 | ] 667 | }, 668 | { 669 | "cell_type": "code", 670 | "execution_count": 21, 671 | "id": "6051b551", 672 | "metadata": {}, 673 | "outputs": [ 674 | { 675 | "data": { 676 | "text/plain": [ 677 | "(40, 5)" 678 | ] 679 | }, 680 | "execution_count": 21, 681 | "metadata": {}, 682 | "output_type": "execute_result" 683 | } 684 | ], 685 | "source": [ 686 | "_temp = [delta, theta, alpha, beta, gamma]\n", 687 | "all_bands = np.hstack(_temp)\n", 688 | "all_bands.shape" 689 | ] 690 | }, 691 | { 692 | "cell_type": "markdown", 693 | "id": "a345bbf9", 694 | "metadata": {}, 695 | "source": [ 696 | "Now let's plot it!" 697 | ] 698 | }, 699 | { 700 | "cell_type": "code", 701 | "execution_count": 22, 702 | "id": "22cdb1db", 703 | "metadata": {}, 704 | "outputs": [ 705 | { 706 | "data": { 707 | "text/html": [ 708 | "
\n", 709 | "\n", 722 | "\n", 723 | " \n", 724 | " \n", 725 | " \n", 726 | " \n", 727 | " \n", 728 | " \n", 729 | " \n", 730 | " \n", 731 | " \n", 732 | " \n", 733 | " \n", 734 | " \n", 735 | " \n", 736 | " \n", 737 | " \n", 738 | " \n", 739 | " \n", 740 | " \n", 741 | " \n", 742 | " \n", 743 | " \n", 744 | " \n", 745 | " \n", 746 | " \n", 747 | " \n", 748 | " \n", 749 | " \n", 750 | " \n", 751 | " \n", 752 | " \n", 753 | " \n", 754 | " \n", 755 | " \n", 756 | " \n", 757 | " \n", 758 | " \n", 759 | " \n", 760 | " \n", 761 | " \n", 762 | " \n", 763 | " \n", 764 | " \n", 765 | " \n", 766 | " \n", 767 | " \n", 768 | " \n", 769 | " \n", 770 | " \n", 771 | " \n", 772 | " \n", 773 | " \n", 774 | " \n", 775 | "
deltathetaalphabetagamma
0-19.416558-0.385428-0.287428-6.156266-31.307289
1-18.5747230.7958950.153971-6.109330-30.781067
2-19.4165560.9299370.823071-5.387953-30.326138
3-19.4384991.1350040.137797-6.246238-30.726357
4-21.480252-2.195351-0.587706-6.380772-31.869710
\n", 776 | "
" 777 | ], 778 | "text/plain": [ 779 | " delta theta alpha beta gamma\n", 780 | "0 -19.416558 -0.385428 -0.287428 -6.156266 -31.307289\n", 781 | "1 -18.574723 0.795895 0.153971 -6.109330 -30.781067\n", 782 | "2 -19.416556 0.929937 0.823071 -5.387953 -30.326138\n", 783 | "3 -19.438499 1.135004 0.137797 -6.246238 -30.726357\n", 784 | "4 -21.480252 -2.195351 -0.587706 -6.380772 -31.869710" 785 | ] 786 | }, 787 | "execution_count": 22, 788 | "metadata": {}, 789 | "output_type": "execute_result" 790 | } 791 | ], 792 | "source": [ 793 | "#first turn it into dataframe\n", 794 | "df = pd.DataFrame(all_bands, columns = ['delta','theta','alpha', 'beta', 'gamma'])\n", 795 | "df.head()" 796 | ] 797 | }, 798 | { 799 | "cell_type": "code", 800 | "execution_count": 23, 801 | "id": "0a254e1b", 802 | "metadata": {}, 803 | "outputs": [ 804 | { 805 | "data": { 806 | "image/png": "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", 807 | "text/plain": [ 808 | "
" 809 | ] 810 | }, 811 | "metadata": { 812 | "needs_background": "light" 813 | }, 814 | "output_type": "display_data" 815 | } 816 | ], 817 | "source": [ 818 | "sns.boxplot(data = df)\n", 819 | "plt.show()" 820 | ] 821 | }, 822 | { 823 | "cell_type": "markdown", 824 | "id": "d86cfa4c", 825 | "metadata": {}, 826 | "source": [ 827 | "Here, it's quite difficult to infer anything, so let's try to see whether there is any differences in these bands across the two valence conditions" 828 | ] 829 | }, 830 | { 831 | "cell_type": "code", 832 | "execution_count": 24, 833 | "id": "c922d274", 834 | "metadata": {}, 835 | "outputs": [ 836 | { 837 | "data": { 838 | "text/plain": [ 839 | "((21, 5), (19, 5))" 840 | ] 841 | }, 842 | "execution_count": 24, 843 | "metadata": {}, 844 | "output_type": "execute_result" 845 | } 846 | ], 847 | "source": [ 848 | "low_valence_cond = (label==0).reshape(-1) #reshape to (40, ) instead of (40, 1)\n", 849 | "high_valence_cond = (label==1).reshape(-1) #reshape to (40, ) instead of (40, 1)\n", 850 | "low_valence = all_bands[low_valence_cond]\n", 851 | "high_valence = all_bands[high_valence_cond]\n", 852 | "low_valence.shape, high_valence.shape" 853 | ] 854 | }, 855 | { 856 | "cell_type": "code", 857 | "execution_count": 25, 858 | "id": "18e1d6c1", 859 | "metadata": {}, 860 | "outputs": [ 861 | { 862 | "data": { 863 | "image/png": "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", 864 | "text/plain": [ 865 | "
" 866 | ] 867 | }, 868 | "metadata": { 869 | "needs_background": "light" 870 | }, 871 | "output_type": "display_data" 872 | } 873 | ], 874 | "source": [ 875 | "sns.boxplot(data = low_valence)\n", 876 | "plt.show()" 877 | ] 878 | }, 879 | { 880 | "cell_type": "code", 881 | "execution_count": 26, 882 | "id": "ee77036e", 883 | "metadata": {}, 884 | "outputs": [ 885 | { 886 | "data": { 887 | "image/png": "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", 888 | "text/plain": [ 889 | "
" 890 | ] 891 | }, 892 | "metadata": { 893 | "needs_background": "light" 894 | }, 895 | "output_type": "display_data" 896 | } 897 | ], 898 | "source": [ 899 | "sns.boxplot(data = high_valence)\n", 900 | "plt.show()" 901 | ] 902 | }, 903 | { 904 | "cell_type": "markdown", 905 | "id": "cf2c33d1", 906 | "metadata": {}, 907 | "source": [ 908 | "Nothing! Almost invisible difference. From this EDA, we can kinda conclude that it is quite pointless to try machine learning here, since we cannot see any difference" 909 | ] 910 | }, 911 | { 912 | "cell_type": "markdown", 913 | "id": "1d955ea7", 914 | "metadata": {}, 915 | "source": [ 916 | "## 5. EDA: Asymmetry\n", 917 | "\n", 918 | "Asymmetry analysis here refers to the analysis of imbalance between left and right symmetrical location. Asymmetry analysis is another very basic and must-do analysis for emotions/cognitions/resting state." 919 | ] 920 | }, 921 | { 922 | "cell_type": "markdown", 923 | "id": "fb9e7c00", 924 | "metadata": {}, 925 | "source": [ 926 | "Here is the channels and their index.\n", 927 | "\n", 928 | "- Channels: 32\n", 929 | " 1.\tFp1\t\n", 930 | " 2.\tAF3\n", 931 | " 3.\tF3\n", 932 | " 4.\tF7\n", 933 | " 5.\tFC5\n", 934 | " 6.\tFC1\n", 935 | " 7.\tC3\n", 936 | " 8.\tT7\n", 937 | " 9.\tCP5\n", 938 | " 10.\tCP1\n", 939 | " 11.\tP3\n", 940 | " 12.\tP7\n", 941 | " 13.\tPO3\n", 942 | " 14.\tO1\n", 943 | " 15.\tOz\n", 944 | " 16.\tPz\n", 945 | " 17.\tFp2\n", 946 | " 18.\tAF4\n", 947 | " 19.\tFz\n", 948 | " 20.\tF4\n", 949 | " 21.\tF8\n", 950 | " 22.\tFC6\n", 951 | " 23.\tFC2\n", 952 | " 24.\tCz\n", 953 | " 25.\tC4\n", 954 | " 26.\tT8\n", 955 | " 27.\tCP6\n", 956 | " 28.\tCP2\n", 957 | " 29.\tP4\n", 958 | " 30.\tP8\n", 959 | " 31.\tPO4\n", 960 | " 32.\tO2\n", 961 | "\n", 962 | "\n", 963 | "\n", 964 | "We will seperate the channels into left-ones and right-ones" 965 | ] 966 | }, 967 | { 968 | "cell_type": "markdown", 969 | "id": "d287455a", 970 | "metadata": {}, 971 | "source": [ 972 | "Based on 10-20 system, the followed number if odd, is on the left (if even, on the right). `z` is middle.\n", 973 | "\n", 974 | "Based on this paper, [EEG-Based Emotion Recognition Using Logistic Regression with Gaussian Kernel and Laplacian Prior and Investigation of Critical Frequency Bands](https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&ved=2ahUKEwj7kei-1Mr2AhVTzDgGHdH_D3oQFnoECAkQAQ&url=https%3A%2F%2Fwww.mdpi.com%2F2076-3417%2F10%2F5%2F1619%2Fpdf&usg=AOvVaw3nwBT1NPFALmqKKm4rbIuE), there are 14 pairs of left-right of asymmetry we can engineer.\n", 975 | "\n", 976 | "1. Fp1-Fp2\n", 977 | "2. F7-F8\n", 978 | "3. F3-F4\n", 979 | "4. T7-T8\n", 980 | "5. P7-P8\n", 981 | "6. C3-C4\n", 982 | "7. P3-P4\n", 983 | "8. O1-O2\n", 984 | "9. AF3-AF4\n", 985 | "10. FC5-FC6\n", 986 | "11. FC1-FC2\n", 987 | "12. CP5-CP6\n", 988 | "13. CP1-CP2\n", 989 | "14. PO3-PO4\n", 990 | "\n", 991 | "And there are two relation we can create. Differential Asymmetry (DASM) and Rational Asymmetry (RASM)\n", 992 | "\n", 993 | "Another 11 pairs of frontal-posterior is as follow\n", 994 | "\n", 995 | "1. FC5-CP5\n", 996 | "2. FC1-CP1\n", 997 | "3. FC2-CP2\n", 998 | "4. FC6-CP6\n", 999 | "5. F7-P7\n", 1000 | "6. F3-P3\n", 1001 | "7. Fz-Pz\n", 1002 | "8. F4-P4\n", 1003 | "9. F8-P8\n", 1004 | "10. Fp1-O1\n", 1005 | "11. Fp2-O2\n", 1006 | "\n", 1007 | "The paper uses differential caudality (DCAU)\n", 1008 | "\n", 1009 | "$$ DASM = DE(X_{left}) - DE(X_{right}) $$\n", 1010 | "$$ RASM = \\frac{DE(X_{left})}{DE(X_{right})} $$\n", 1011 | "$$ DCAU = DE(X_{frontal}) - DE(X_{posterior}) $$\n", 1012 | "\n", 1013 | "$ DE() $ is a function that convert PSD to log-PSD which we already did." 1014 | ] 1015 | }, 1016 | { 1017 | "cell_type": "code", 1018 | "execution_count": 27, 1019 | "id": "ba82ff50", 1020 | "metadata": {}, 1021 | "outputs": [ 1022 | { 1023 | "name": "stdout", 1024 | "output_type": "stream", 1025 | "text": [ 1026 | "left_channel_indexes=[0, 3, 2, 7, 11, 6, 10, 13, 1, 4, 5, 8, 9, 12]\n", 1027 | "right_channel_indexes=[16, 20, 19, 25, 29, 24, 28, 31, 17, 21, 22, 26, 27, 30]\n", 1028 | "frontal_channel_indexes=[4, 5, 22, 21, 3, 2, 18, 19, 20, 0, 16]\n", 1029 | "posterior_channel_indexes=[8, 9, 27, 26, 11, 10, 15, 28, 29, 13, 31]\n" 1030 | ] 1031 | } 1032 | ], 1033 | "source": [ 1034 | "channels = ['Fp1','AF3','F3','F7','FC5','FC1','C3','T7','CP5','CP1','P3','P7','PO3','O1','Oz','Pz','Fp2','AF4','Fz','F4','F8','FC6','FC2','Cz','C4','T8','CP6','CP2','P4','P8','PO4','O2']\n", 1035 | "left_channels = ['Fp1','F7','F3','T7','P7','C3','P3','O1','AF3','FC5','FC1','CP5','CP1','PO3']\n", 1036 | "right_channels = ['Fp2','F8','F4','T8','P8','C4','P4','O2','AF4','FC6','FC2','CP6','CP2','PO4']\n", 1037 | "left_channel_indexes = [ channels.index(ch) for ch in left_channels ]\n", 1038 | "right_channel_indexes = [ channels.index(ch) for ch in right_channels ]\n", 1039 | "\n", 1040 | "print(f\"{left_channel_indexes=}\")\n", 1041 | "print(f\"{right_channel_indexes=}\")\n", 1042 | "\n", 1043 | "frontal_channels = ['FC5','FC1','FC2','FC6','F7','F3','Fz','F4','F8','Fp1','Fp2']\n", 1044 | "posterior_channels = ['CP5','CP1','CP2','CP6','P7','P3','Pz','P4','P8','O1','O2']\n", 1045 | "\n", 1046 | "frontal_channel_indexes = [ channels.index(ch) for ch in frontal_channels ]\n", 1047 | "posterior_channel_indexes = [ channels.index(ch) for ch in posterior_channels ]\n", 1048 | "\n", 1049 | "print(f\"{frontal_channel_indexes=}\")\n", 1050 | "print(f\"{posterior_channel_indexes=}\")" 1051 | ] 1052 | }, 1053 | { 1054 | "cell_type": "markdown", 1055 | "id": "2f766d2c", 1056 | "metadata": {}, 1057 | "source": [ 1058 | "Let's get PSD. I will use `MNE_feature` since it handles epochs-like data for me and I don't need to analyze the data (remember, the data is already preprocessed)\n", 1059 | "\n", 1060 | "The idea is as follow;\n", 1061 | "\n", 1062 | "1. I will slowly calculate PSD of each frequncy band.\n", 1063 | "2. For each band, I will calculate DASM, RASM, and DCAU" 1064 | ] 1065 | }, 1066 | { 1067 | "cell_type": "code", 1068 | "execution_count": 28, 1069 | "id": "79f316b7", 1070 | "metadata": {}, 1071 | "outputs": [ 1072 | { 1073 | "name": "stdout", 1074 | "output_type": "stream", 1075 | "text": [ 1076 | "________________________________________________________________________________\n", 1077 | "[Memory] Calling mne_features.feature_extraction.extract_features...\n", 1078 | "extract_features(tensor([[[ 0.9482, 1.6533, 3.0137, ..., -2.8265, -4.4772, -3.6769],\n", 1079 | " [ 0.1247, 1.3901, 1.8351, ..., -2.9870, -6.2878, -4.4743],\n", 1080 | " [ -2.2165, 2.2920, 2.7464, ..., -2.6371, -7.4065, -6.7559],\n", 1081 | " ...,\n", 1082 | " [ -3.5572, -1.2603, -3.0996, ..., 6.6889, 6.9123, 5.6690],\n", 1083 | " [ -1.2373, -0.9414, 1.0280, ..., 2.6402, 6.5746, 7.2085],\n", 1084 | " [ 0.3723, 2.0762, 4.4652, ..., 2.2396, 3.1826, 4.7081]],\n", 1085 | "\n", 1086 | " [[ 10.2602, 12.7954, 10.4262, ..., 6.0222, 7.5391, 9.3522],\n", 1087 | " [ 9.4919, 12.5898, 10.5740, ..., 6.0340, 9.0687, 8.7402],\n", 1088 | " [ 7.1287, 12.2065, 9.4965, ..., 6.1797, 6.9337, ..., \n", 1089 | "128, ['pow_freq_bands'], funcs_params={ 'pow_freq_bands__freq_bands': (8, 12),\n", 1090 | " 'pow_freq_bands__log': True,\n", 1091 | " 'pow_freq_bands__normalize': False}, n_jobs=8)\n", 1092 | "_______________________________________________extract_features - 102.0s, 1.7min\n", 1093 | "(40, 32) (40, 14) (40, 14) (40, 11)\n" 1094 | ] 1095 | } 1096 | ], 1097 | "source": [ 1098 | "from mne_features.feature_extraction import FeatureExtractor\n", 1099 | "\n", 1100 | "# bands = dict({\n", 1101 | "# \"delta\": (0,4), \n", 1102 | "# \"theta\": (4,8),\n", 1103 | "# \"alpha\": (8,12),\n", 1104 | "# \"beta\": (12,30),\n", 1105 | "# \"gamma\": (30,64)\n", 1106 | "# })\n", 1107 | "\n", 1108 | "#for simplicity, we gonna just try alpha asymmetry\n", 1109 | "bands = dict({\n", 1110 | " \"alpha\": (8,12),\n", 1111 | "})\n", 1112 | "\n", 1113 | "PSDs = []\n", 1114 | "DASMs = []\n", 1115 | "RASMs = []\n", 1116 | "DCAUs = []\n", 1117 | "for band_name,band_range in bands.items():\n", 1118 | " params = dict({\n", 1119 | " 'pow_freq_bands__log':True,\n", 1120 | " 'pow_freq_bands__normalize':False,\n", 1121 | " 'pow_freq_bands__freq_bands':band_range\n", 1122 | " })\n", 1123 | " fe = FeatureExtractor(sfreq=128, selected_funcs=['pow_freq_bands'],params=params,n_jobs=8, memory=\"cache/\")\n", 1124 | " PSD = fe.fit_transform(X=data)\n", 1125 | " # (40 samples, 32 channels)\n", 1126 | " # print(\"1\", PSD.shape)\n", 1127 | "\n", 1128 | " PSD_left = PSD[:, left_channel_indexes].copy()\n", 1129 | " PSD_right = PSD[:, right_channel_indexes].copy()\n", 1130 | " PSD_frontal = PSD[:, frontal_channel_indexes].copy()\n", 1131 | " PSD_posterior = PSD[:, posterior_channel_indexes].copy()\n", 1132 | " # (40, 14) (40, 14) (40, 11) (40, 11)\n", 1133 | " # print(\"2\", PSD_left.shape, PSD_right.shape, PSD_frontal.shape, PSD_posterior.shape)\n", 1134 | " DASM = PSD_left - PSD_right\n", 1135 | " # (40, 14)\n", 1136 | " # print(\"3\", DASM.shape)\n", 1137 | " RASM = PSD_left / PSD_right\n", 1138 | " # (40, 14)\n", 1139 | " # print(\"4\", RASM.shape)\n", 1140 | " DCAU = PSD_frontal - PSD_posterior\n", 1141 | " # (40, 11)\n", 1142 | " # print(\"5\", DCAU.shape)\n", 1143 | "\n", 1144 | " # Will be use later for comparison\n", 1145 | " PSDs.append(PSD)\n", 1146 | " DASMs.append(DASM)\n", 1147 | " RASMs.append(RASM)\n", 1148 | " DCAUs.append(DCAU)\n", 1149 | "\n", 1150 | "PSDs = np.hstack(PSDs)\n", 1151 | "DASMs = np.hstack(DASMs)\n", 1152 | "RASMs = np.hstack(RASMs)\n", 1153 | "DCAUs = np.hstack(DCAUs)\n", 1154 | "\n", 1155 | "# (40, 32) (40, 14) (40, 14) (40, 11)\n", 1156 | "print(PSDs.shape, DASMs.shape, RASMs.shape, DCAUs.shape)" 1157 | ] 1158 | }, 1159 | { 1160 | "cell_type": "markdown", 1161 | "id": "0c4789dc", 1162 | "metadata": {}, 1163 | "source": [ 1164 | "We can try to first look at DASM" 1165 | ] 1166 | }, 1167 | { 1168 | "cell_type": "code", 1169 | "execution_count": 29, 1170 | "id": "c922d274", 1171 | "metadata": {}, 1172 | "outputs": [ 1173 | { 1174 | "data": { 1175 | "text/plain": [ 1176 | "((21, 14), (19, 14))" 1177 | ] 1178 | }, 1179 | "execution_count": 29, 1180 | "metadata": {}, 1181 | "output_type": "execute_result" 1182 | } 1183 | ], 1184 | "source": [ 1185 | "low_valence = DASM[low_valence_cond]\n", 1186 | "high_valence = DASM[high_valence_cond]\n", 1187 | "low_valence.shape, high_valence.shape" 1188 | ] 1189 | }, 1190 | { 1191 | "cell_type": "code", 1192 | "execution_count": 30, 1193 | "id": "18e1d6c1", 1194 | "metadata": {}, 1195 | "outputs": [ 1196 | { 1197 | "data": { 1198 | "text/plain": [ 1199 | "(40, 15)" 1200 | ] 1201 | }, 1202 | "execution_count": 30, 1203 | "metadata": {}, 1204 | "output_type": "execute_result" 1205 | } 1206 | ], 1207 | "source": [ 1208 | "dcau_channels = ['FC5-CP5','FC2-CP2','FC6-CP6','F7-P7','F3-P3','Fz-Pz','F4-P4',\n", 1209 | " 'F8-P8','FC1-CP1','Fp1-O1','Fp2-O2']\n", 1210 | "dasm_channels = ['Fp1-Fp2', 'F7-F8','F3-F4', 'T7-T8','P7-P8','C3-C4','P3-P4',\n", 1211 | " 'O1-O2','AF3-AF4','FC5-FC6','FC1-FC2','CP5-CP6','CP1-CP2','PO3-PO4']\n", 1212 | "\n", 1213 | "low_valence_df = pd.DataFrame(low_valence, columns = dasm_channels)\n", 1214 | "low_valence_df['valence'] = 'low'\n", 1215 | "\n", 1216 | "high_valence_df = pd.DataFrame(high_valence, columns = dasm_channels)\n", 1217 | "high_valence_df['valence'] = 'high'\n", 1218 | "\n", 1219 | "df_concat = pd.concat([low_valence_df, high_valence_df], axis=0)\n", 1220 | "\n", 1221 | "df_concat.shape" 1222 | ] 1223 | }, 1224 | { 1225 | "cell_type": "code", 1226 | "execution_count": 31, 1227 | "id": "d18e3004", 1228 | "metadata": {}, 1229 | "outputs": [ 1230 | { 1231 | "data": { 1232 | "text/html": [ 1233 | "
\n", 1234 | "\n", 1247 | "\n", 1248 | " \n", 1249 | " \n", 1250 | " \n", 1251 | " \n", 1252 | " \n", 1253 | " \n", 1254 | " \n", 1255 | " \n", 1256 | " \n", 1257 | " \n", 1258 | " \n", 1259 | " \n", 1260 | " \n", 1261 | " \n", 1262 | " \n", 1263 | " \n", 1264 | " \n", 1265 | " \n", 1266 | " \n", 1267 | " \n", 1268 | " \n", 1269 | " \n", 1270 | " \n", 1271 | " \n", 1272 | " \n", 1273 | " \n", 1274 | " \n", 1275 | " \n", 1276 | " \n", 1277 | " \n", 1278 | " \n", 1279 | " \n", 1280 | " \n", 1281 | " \n", 1282 | " \n", 1283 | " \n", 1284 | " \n", 1285 | " \n", 1286 | " \n", 1287 | " \n", 1288 | " \n", 1289 | " \n", 1290 | " \n", 1291 | " \n", 1292 | " \n", 1293 | " \n", 1294 | " \n", 1295 | " \n", 1296 | " \n", 1297 | " \n", 1298 | " \n", 1299 | " \n", 1300 | " \n", 1301 | " \n", 1302 | " \n", 1303 | " \n", 1304 | " \n", 1305 | " \n", 1306 | " \n", 1307 | " \n", 1308 | " \n", 1309 | " \n", 1310 | " \n", 1311 | " \n", 1312 | " \n", 1313 | " \n", 1314 | " \n", 1315 | " \n", 1316 | " \n", 1317 | " \n", 1318 | " \n", 1319 | " \n", 1320 | " \n", 1321 | " \n", 1322 | " \n", 1323 | " \n", 1324 | " \n", 1325 | " \n", 1326 | " \n", 1327 | " \n", 1328 | " \n", 1329 | " \n", 1330 | " \n", 1331 | " \n", 1332 | " \n", 1333 | " \n", 1334 | " \n", 1335 | " \n", 1336 | " \n", 1337 | " \n", 1338 | " \n", 1339 | " \n", 1340 | " \n", 1341 | " \n", 1342 | " \n", 1343 | " \n", 1344 | " \n", 1345 | " \n", 1346 | " \n", 1347 | " \n", 1348 | " \n", 1349 | " \n", 1350 | " \n", 1351 | " \n", 1352 | " \n", 1353 | " \n", 1354 | " \n", 1355 | " \n", 1356 | " \n", 1357 | " \n", 1358 | " \n", 1359 | " \n", 1360 | "
Fp1-Fp2F7-F8F3-F4T7-T8P7-P8C3-C4P3-P4O1-O2AF3-AF4FC5-FC6FC1-FC2CP5-CP6CP1-CP2PO3-PO4valence
0-0.0620190.038666-0.0601960.1029280.268888-0.0834800.001628-0.068236-0.007693-0.089520-0.2329390.133538-0.0538380.028254low
1-0.1108740.140154-0.1295250.1261450.275887-0.1050840.0182080.023864-0.060906-0.075915-0.2333540.189782-0.0005450.078809low
2-0.0501650.171666-0.0130330.1672440.169724-0.0936710.083920-0.0255550.0608510.029082-0.2546860.2356390.0811340.158544low
3-0.0859220.056472-0.0752950.1017010.2598930.0002860.1868390.009887-0.048551-0.080910-0.1392790.2737220.1759950.185218low
4-0.0516580.055029-0.0391490.0375200.281425-0.0123690.0917160.0342320.040922-0.069554-0.2213180.1949740.0864630.224313low
\n", 1361 | "
" 1362 | ], 1363 | "text/plain": [ 1364 | " Fp1-Fp2 F7-F8 F3-F4 T7-T8 P7-P8 C3-C4 P3-P4 \\\n", 1365 | "0 -0.062019 0.038666 -0.060196 0.102928 0.268888 -0.083480 0.001628 \n", 1366 | "1 -0.110874 0.140154 -0.129525 0.126145 0.275887 -0.105084 0.018208 \n", 1367 | "2 -0.050165 0.171666 -0.013033 0.167244 0.169724 -0.093671 0.083920 \n", 1368 | "3 -0.085922 0.056472 -0.075295 0.101701 0.259893 0.000286 0.186839 \n", 1369 | "4 -0.051658 0.055029 -0.039149 0.037520 0.281425 -0.012369 0.091716 \n", 1370 | "\n", 1371 | " O1-O2 AF3-AF4 FC5-FC6 FC1-FC2 CP5-CP6 CP1-CP2 PO3-PO4 \\\n", 1372 | "0 -0.068236 -0.007693 -0.089520 -0.232939 0.133538 -0.053838 0.028254 \n", 1373 | "1 0.023864 -0.060906 -0.075915 -0.233354 0.189782 -0.000545 0.078809 \n", 1374 | "2 -0.025555 0.060851 0.029082 -0.254686 0.235639 0.081134 0.158544 \n", 1375 | "3 0.009887 -0.048551 -0.080910 -0.139279 0.273722 0.175995 0.185218 \n", 1376 | "4 0.034232 0.040922 -0.069554 -0.221318 0.194974 0.086463 0.224313 \n", 1377 | "\n", 1378 | " valence \n", 1379 | "0 low \n", 1380 | "1 low \n", 1381 | "2 low \n", 1382 | "3 low \n", 1383 | "4 low " 1384 | ] 1385 | }, 1386 | "execution_count": 31, 1387 | "metadata": {}, 1388 | "output_type": "execute_result" 1389 | } 1390 | ], 1391 | "source": [ 1392 | "df_concat.head()" 1393 | ] 1394 | }, 1395 | { 1396 | "cell_type": "code", 1397 | "execution_count": 32, 1398 | "id": "ee77036e", 1399 | "metadata": {}, 1400 | "outputs": [ 1401 | { 1402 | "data": { 1403 | "image/png": "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", 1404 | "text/plain": [ 1405 | "
" 1406 | ] 1407 | }, 1408 | "metadata": { 1409 | "needs_background": "light" 1410 | }, 1411 | "output_type": "display_data" 1412 | }, 1413 | { 1414 | "data": { 1415 | "image/png": "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", 1416 | "text/plain": [ 1417 | "
" 1418 | ] 1419 | }, 1420 | "metadata": { 1421 | "needs_background": "light" 1422 | }, 1423 | "output_type": "display_data" 1424 | }, 1425 | { 1426 | "data": { 1427 | "image/png": "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", 1428 | "text/plain": [ 1429 | "
" 1430 | ] 1431 | }, 1432 | "metadata": { 1433 | "needs_background": "light" 1434 | }, 1435 | "output_type": "display_data" 1436 | }, 1437 | { 1438 | "data": { 1439 | "image/png": "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", 1440 | "text/plain": [ 1441 | "
" 1442 | ] 1443 | }, 1444 | "metadata": { 1445 | "needs_background": "light" 1446 | }, 1447 | "output_type": "display_data" 1448 | }, 1449 | { 1450 | "data": { 1451 | "image/png": "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", 1452 | "text/plain": [ 1453 | "
" 1454 | ] 1455 | }, 1456 | "metadata": { 1457 | "needs_background": "light" 1458 | }, 1459 | "output_type": "display_data" 1460 | }, 1461 | { 1462 | "data": { 1463 | "image/png": "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", 1464 | "text/plain": [ 1465 | "
" 1466 | ] 1467 | }, 1468 | "metadata": { 1469 | "needs_background": "light" 1470 | }, 1471 | "output_type": "display_data" 1472 | }, 1473 | { 1474 | "data": { 1475 | "image/png": "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", 1476 | "text/plain": [ 1477 | "
" 1478 | ] 1479 | }, 1480 | "metadata": { 1481 | "needs_background": "light" 1482 | }, 1483 | "output_type": "display_data" 1484 | }, 1485 | { 1486 | "data": { 1487 | "image/png": "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", 1488 | "text/plain": [ 1489 | "
" 1490 | ] 1491 | }, 1492 | "metadata": { 1493 | "needs_background": "light" 1494 | }, 1495 | "output_type": "display_data" 1496 | }, 1497 | { 1498 | "data": { 1499 | "image/png": "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", 1500 | "text/plain": [ 1501 | "
" 1502 | ] 1503 | }, 1504 | "metadata": { 1505 | "needs_background": "light" 1506 | }, 1507 | "output_type": "display_data" 1508 | }, 1509 | { 1510 | "data": { 1511 | "image/png": "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", 1512 | "text/plain": [ 1513 | "
" 1514 | ] 1515 | }, 1516 | "metadata": { 1517 | "needs_background": "light" 1518 | }, 1519 | "output_type": "display_data" 1520 | }, 1521 | { 1522 | "data": { 1523 | "image/png": "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", 1524 | "text/plain": [ 1525 | "
" 1526 | ] 1527 | }, 1528 | "metadata": { 1529 | "needs_background": "light" 1530 | }, 1531 | "output_type": "display_data" 1532 | }, 1533 | { 1534 | "data": { 1535 | "image/png": "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", 1536 | "text/plain": [ 1537 | "
" 1538 | ] 1539 | }, 1540 | "metadata": { 1541 | "needs_background": "light" 1542 | }, 1543 | "output_type": "display_data" 1544 | }, 1545 | { 1546 | "data": { 1547 | "image/png": "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", 1548 | "text/plain": [ 1549 | "
" 1550 | ] 1551 | }, 1552 | "metadata": { 1553 | "needs_background": "light" 1554 | }, 1555 | "output_type": "display_data" 1556 | }, 1557 | { 1558 | "data": { 1559 | "image/png": "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", 1560 | "text/plain": [ 1561 | "
" 1562 | ] 1563 | }, 1564 | "metadata": { 1565 | "needs_background": "light" 1566 | }, 1567 | "output_type": "display_data" 1568 | } 1569 | ], 1570 | "source": [ 1571 | "for col in dasm_channels:\n", 1572 | " sns.boxplot(data = df_concat, x = 'valence', y = col)\n", 1573 | " plt.show() " 1574 | ] 1575 | }, 1576 | { 1577 | "cell_type": "markdown", 1578 | "id": "13f3f5fd", 1579 | "metadata": {}, 1580 | "source": [ 1581 | "Nice! after scanning, we can see some very observable difference, especially in the frontal nodes. This proves that asymmetry is kinda useful to understand emotion! What you can do next is probably trying other asymmetry features, as well as to input strong features into some machine learning algorithm. You can even try deep learning algorithm!" 1582 | ] 1583 | } 1584 | ], 1585 | "metadata": { 1586 | "kernelspec": { 1587 | "display_name": "Python 3.8.6 ('teaching_env')", 1588 | "language": "python", 1589 | "name": "python3" 1590 | }, 1591 | "language_info": { 1592 | "codemirror_mode": { 1593 | "name": "ipython", 1594 | "version": 3 1595 | }, 1596 | "file_extension": ".py", 1597 | "mimetype": "text/x-python", 1598 | "name": "python", 1599 | "nbconvert_exporter": "python", 1600 | "pygments_lexer": "ipython3", 1601 | "version": "3.8.6" 1602 | }, 1603 | "vscode": { 1604 | "interpreter": { 1605 | "hash": "becc4c8e5ad229b2591d820334d85e3db0111492344629bf57f272470dce75a5" 1606 | } 1607 | } 1608 | }, 1609 | "nbformat": 4, 1610 | "nbformat_minor": 5 1611 | } 1612 | --------------------------------------------------------------------------------