├── src ├── engine.py ├── loss.py ├── utils.py ├── __init__.py ├── dataset.py ├── metrics.py ├── feature_generator.py ├── dispatcher.py ├── create_folds.py ├── predict.py └── train.py ├── run.sh └── .gitignore /src/engine.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /src/loss.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /src/utils.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /src/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /src/dataset.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /src/metrics.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /src/feature_generator.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /src/dispatcher.py: -------------------------------------------------------------------------------- 1 | from sklearn import ensemble 2 | 0.75091 3 | MODELS = { 4 | "randomforest": ensemble.RandomForestClassifier(n_estimators=200, n_jobs=-1, verbose=2), 5 | "extratrees": ensemble.ExtraTreesClassifier(n_estimators=200, n_jobs=-1, verbose=2), 6 | } -------------------------------------------------------------------------------- /run.sh: -------------------------------------------------------------------------------- 1 | export TRAINING_DATA=input/train_folds.csv 2 | export TEST_DATA=input/test.csv 3 | 4 | export MODEL=$1 5 | 6 | #FOLD=0 python -m src.train 7 | #FOLD=1 python -m src.train 8 | #FOLD=2 python -m src.train 9 | #FOLD=3 python -m src.train 10 | #FOLD=4 python -m src.train 11 | python -m src.predict -------------------------------------------------------------------------------- /src/create_folds.py: -------------------------------------------------------------------------------- 1 | import pandas as pd 2 | from sklearn import model_selection 3 | 4 | if __name__ == "__main__": 5 | df = pd.read_csv("input/train.csv") 6 | df["kfold"] = -1 7 | 8 | df = df.sample(frac=1).reset_index(drop=True) 9 | 10 | kf = model_selection.StratifiedKFold(n_splits=5, shuffle=False, random_state=42) 11 | 12 | 13 | for fold, (train_idx, val_idx) in enumerate(kf.split(X=df, y=df.target.values)): 14 | print(len(train_idx), len(val_idx)) 15 | df.loc[val_idx, 'kfold'] = fold 16 | 17 | 18 | df.to_csv("input/train_folds.csv", index=False) 19 | -------------------------------------------------------------------------------- /src/predict.py: -------------------------------------------------------------------------------- 1 | import os 2 | import pandas as pd 3 | from sklearn import ensemble 4 | from sklearn import preprocessing 5 | from sklearn import metrics 6 | import joblib 7 | import numpy as np 8 | 9 | from . import dispatcher 10 | 11 | TEST_DATA = os.environ.get("TEST_DATA") 12 | MODEL = os.environ.get("MODEL") 13 | 14 | def predict(): 15 | df = pd.read_csv(TEST_DATA) 16 | test_idx = df["id"].values 17 | predictions = None 18 | 19 | for FOLD in range(5): 20 | print(FOLD) 21 | df = pd.read_csv(TEST_DATA) 22 | encoders = joblib.load(os.path.join("models", f"{MODEL}_{FOLD}_label_encoder.pkl")) 23 | cols = joblib.load(os.path.join("models", f"{MODEL}_{FOLD}_columns.pkl")) 24 | for c in encoders: 25 | print(c) 26 | lbl = encoders[c] 27 | df.loc[:, c] = lbl.transform(df[c].values.tolist()) 28 | 29 | # data is ready to train 30 | clf = joblib.load(os.path.join("models", f"{MODEL}_{FOLD}.pkl")) 31 | 32 | df = df[cols] 33 | preds = clf.predict_proba(df)[:, 1] 34 | 35 | if FOLD == 0: 36 | predictions = preds 37 | else: 38 | predictions += preds 39 | 40 | predictions /= 5 41 | 42 | sub = pd.DataFrame(np.column_stack((test_idx, predictions)), columns=["id", "target"]) 43 | return sub 44 | 45 | 46 | if __name__ == "__main__": 47 | submission = predict() 48 | submission.to_csv(f"models/{MODEL}.csv", index=False) 49 | -------------------------------------------------------------------------------- /src/train.py: -------------------------------------------------------------------------------- 1 | import os 2 | import pandas as pd 3 | from sklearn import ensemble 4 | from sklearn import preprocessing 5 | from sklearn import metrics 6 | import joblib 7 | 8 | from . import dispatcher 9 | 10 | TRAINING_DATA = os.environ.get("TRAINING_DATA") 11 | TEST_DATA = os.environ.get("TEST_DATA") 12 | FOLD = int(os.environ.get("FOLD")) 13 | MODEL = os.environ.get("MODEL") 14 | 15 | FOLD_MAPPPING = { 16 | 0: [1, 2, 3, 4], 17 | 1: [0, 2, 3, 4], 18 | 2: [0, 1, 3, 4], 19 | 3: [0, 1, 2, 4], 20 | 4: [0, 1, 2, 3] 21 | } 22 | 23 | if __name__ == "__main__": 24 | df = pd.read_csv(TRAINING_DATA) 25 | df_test = pd.read_csv(TEST_DATA) 26 | train_df = df[df.kfold.isin(FOLD_MAPPPING.get(FOLD))].reset_index(drop=True) 27 | valid_df = df[df.kfold==FOLD].reset_index(drop=True) 28 | 29 | ytrain = train_df.target.values 30 | yvalid = valid_df.target.values 31 | 32 | train_df = train_df.drop(["id", "target", "kfold"], axis=1) 33 | valid_df = valid_df.drop(["id", "target", "kfold"], axis=1) 34 | 35 | valid_df = valid_df[train_df.columns] 36 | 37 | label_encoders = {} 38 | for c in train_df.columns: 39 | lbl = preprocessing.LabelEncoder() 40 | lbl.fit(train_df[c].values.tolist() + valid_df[c].values.tolist() + df_test[c].values.tolist()) 41 | train_df.loc[:, c] = lbl.transform(train_df[c].values.tolist()) 42 | valid_df.loc[:, c] = lbl.transform(valid_df[c].values.tolist()) 43 | label_encoders[c] = lbl 44 | 45 | # data is ready to train 46 | clf = dispatcher.MODELS[MODEL] 47 | clf.fit(train_df, ytrain) 48 | preds = clf.predict_proba(valid_df)[:, 1] 49 | print(metrics.roc_auc_score(yvalid, preds)) 50 | 51 | joblib.dump(label_encoders, f"models/{MODEL}_{FOLD}_label_encoder.pkl") 52 | joblib.dump(clf, f"models/{MODEL}_{FOLD}.pkl") 53 | joblib.dump(train_df.columns, f"models/{MODEL}_{FOLD}_columns.pkl") 54 | -------------------------------------------------------------------------------- /.gitignore: -------------------------------------------------------------------------------- 1 | # Byte-compiled / optimized / DLL files 2 | __pycache__/ 3 | *.py[cod] 4 | *$py.class 5 | 6 | # C extensions 7 | *.so 8 | 9 | # Distribution / packaging 10 | .Python 11 | build/ 12 | develop-eggs/ 13 | dist/ 14 | downloads/ 15 | eggs/ 16 | .eggs/ 17 | lib/ 18 | lib64/ 19 | parts/ 20 | sdist/ 21 | var/ 22 | wheels/ 23 | pip-wheel-metadata/ 24 | share/python-wheels/ 25 | *.egg-info/ 26 | .installed.cfg 27 | *.egg 28 | MANIFEST 29 | 30 | # PyInstaller 31 | # Usually these files are written by a python script from a template 32 | # before PyInstaller builds the exe, so as to inject date/other infos into it. 33 | *.manifest 34 | *.spec 35 | 36 | # Installer logs 37 | pip-log.txt 38 | pip-delete-this-directory.txt 39 | 40 | # Unit test / coverage reports 41 | htmlcov/ 42 | .tox/ 43 | .nox/ 44 | .coverage 45 | .coverage.* 46 | .cache 47 | nosetests.xml 48 | coverage.xml 49 | *.cover 50 | *.py,cover 51 | .hypothesis/ 52 | .pytest_cache/ 53 | 54 | # Translations 55 | *.mo 56 | *.pot 57 | 58 | # Django stuff: 59 | *.log 60 | local_settings.py 61 | db.sqlite3 62 | db.sqlite3-journal 63 | 64 | # Flask stuff: 65 | instance/ 66 | .webassets-cache 67 | 68 | # Scrapy stuff: 69 | .scrapy 70 | 71 | # Sphinx documentation 72 | docs/_build/ 73 | 74 | # PyBuilder 75 | target/ 76 | 77 | # Jupyter Notebook 78 | .ipynb_checkpoints 79 | 80 | # IPython 81 | profile_default/ 82 | ipython_config.py 83 | 84 | # pyenv 85 | .python-version 86 | 87 | # pipenv 88 | # According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control. 89 | # However, in case of collaboration, if having platform-specific dependencies or dependencies 90 | # having no cross-platform support, pipenv may install dependencies that don't work, or not 91 | # install all needed dependencies. 92 | #Pipfile.lock 93 | 94 | # PEP 582; used by e.g. github.com/David-OConnor/pyflow 95 | __pypackages__/ 96 | 97 | # Celery stuff 98 | celerybeat-schedule 99 | celerybeat.pid 100 | 101 | # SageMath parsed files 102 | *.sage.py 103 | 104 | # Environments 105 | .env 106 | .venv 107 | env/ 108 | venv/ 109 | ENV/ 110 | env.bak/ 111 | venv.bak/ 112 | 113 | # Spyder project settings 114 | .spyderproject 115 | .spyproject 116 | 117 | # Rope project settings 118 | .ropeproject 119 | 120 | # mkdocs documentation 121 | /site 122 | 123 | # mypy 124 | .mypy_cache/ 125 | .dmypy.json 126 | dmypy.json 127 | 128 | # Pyre type checker 129 | .pyre/ 130 | 131 | 132 | # input data and models 133 | input/ 134 | models/ 135 | 136 | 137 | # data files 138 | *.csv 139 | *.h5 140 | *.pkl 141 | *.pth 142 | --------------------------------------------------------------------------------