{"cells":[{"metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true},"cell_type":"code","source":"# This Python 3 environment comes with many helpful analytics libraries installed\n# It is defined by the kaggle/python docker image: https://github.com/kaggle/docker-python\n# For example, here's several helpful packages to load in \n\nimport numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n\n# Input data files are available in the \"../input/\" directory.\n# For example, running this (by clicking run or pressing Shift+Enter) will list the files in the input directory\n\nimport os\nprint(os.listdir(\"../input\"))\n\n# Any results you write to the current directory are saved as output.","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train = pd.read_csv('../input/train.csv')\ntest = pd.read_csv('../input/test.csv')\nstructures = pd.read_csv('../input/structures.csv')\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"def append_structures(df):\n    # JOIN on atom_index_0\n    df = pd.merge(df, structures, left_on=['molecule_name', 'atom_index_0'], right_on=['molecule_name', 'atom_index'])\n    df = df.rename({'x': 'x_0', 'y': 'y_0', 'z': 'z_0', 'atom': 'atom_0'}, axis=1)\n    df = df.drop(['atom_index', 'atom_index_0'], axis=1)\n\n    # JOIN on atom_index_1\n    df = pd.merge(df, structures, left_on=['molecule_name', 'atom_index_1'], right_on=['molecule_name', 'atom_index'])\n    df = df.rename({'x': 'x_1', 'y': 'y_1', 'z': 'z_1', 'atom': 'atom_1'}, axis=1)\n    df = df.drop(['atom_index', 'atom_index_1'], axis=1)\n    \n    return df","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"def add_molecule_features(df):\n    foo = pd.DataFrame({'1JHC_total': X.groupby('molecule_name')['1JHC'].sum()})\n    foo['molecule_name'] = foo.index\n    df = pd.merge(df, foo, on = ['molecule_name'])\n\n    foo = pd.DataFrame({'1JHN_total': X.groupby('molecule_name')['1JHN'].sum()})\n    foo['molecule_name'] = foo.index\n    df = pd.merge(df, foo, on = ['molecule_name'])\n\n    foo = pd.DataFrame({'2JHC_total': X.groupby('molecule_name')['2JHC'].sum()})\n    foo['molecule_name'] = foo.index\n    df = pd.merge(df, foo, on = ['molecule_name'])\n\n    foo = pd.DataFrame({'2JHN_total': X.groupby('molecule_name')['2JHN'].sum()})\n    foo['molecule_name'] = foo.index\n    df = pd.merge(df, foo, on = ['molecule_name'])\n\n    foo = pd.DataFrame({'2JHH_total': X.groupby('molecule_name')['2JHH'].sum()})\n    foo['molecule_name'] = foo.index\n    df = pd.merge(df, foo, on = ['molecule_name'])\n\n    foo = pd.DataFrame({'3JHC_total': X.groupby('molecule_name')['3JHC'].sum()})\n    foo['molecule_name'] = foo.index\n    df = pd.merge(df, foo, on = ['molecule_name'])\n\n    foo = pd.DataFrame({'3JHH_total': X.groupby('molecule_name')['3JHH'].sum()})\n    foo['molecule_name'] = foo.index\n    df = pd.merge(df, foo, on = ['molecule_name'])\n    \n    foo = pd.DataFrame({'3JHN_total': X.groupby('molecule_name')['3JHN'].sum()})\n    foo['molecule_name'] = foo.index\n    df = pd.merge(df, foo, on = ['molecule_name'])\n\n    return df\n    \n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"def set_features(df):\n    df = append_structures(df)\n    \n    ### Feature engineering goes here\n    \n    # For now drop non-numeric\n    df = df.drop(['atom_0'], axis=1)\n    df = df.drop(['atom_1'], axis=1)\n    \n    df = df.drop(['molecule_name'], axis=1)\n    dummies = pd.get_dummies(df['type'])\n    df = pd.concat([df.drop(['type'], axis=1), dummies], axis=1)\n    return df\n    ","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"from sklearn.ensemble import RandomForestRegressor\n\nregr = RandomForestRegressor(max_depth=2, random_state=0, n_estimators=100)\n\nY = train['scalar_coupling_constant']\nX = train.drop(['scalar_coupling_constant'], axis=1)\n\nX = set_features(X)\n\nregr.fit(X.head(10000),Y.head(10000))\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"Y_hat = regr.predict(X)\n\n","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"test_X = test\ntest_X = set_features(test_X)\ntest_Y = regr.predict(test_X)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"submission_df = pd.DataFrame({'id': test_X['id'], 'scalar_coupling_constant': test_Y})\n\nsubmission_df.to_csv('submission.csv', index=False)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"!wc -l submission.csv\n","execution_count":null,"outputs":[]}],"metadata":{"kernelspec":{"display_name":"Python 3","language":"python","name":"python3"},"language_info":{"name":"python","version":"3.6.4","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"}},"nbformat":4,"nbformat_minor":1}