def main(): df, df_store = load_data(debug=False) feat_matrix, features_x, feature_y = extract_features(df, df_store) quick_score(feat_matrix.loc[~(feat_matrix['Type'] == 'test')], features_x, feature_y) ts = datetime.datetime.fromtimestamp( time.time()).strftime('%Y-%m-%d %H:%M:%S') print(ts)
def main(): df, df_store = load_data(debug=False) feat_matrix, features_x, feature_y = extract_features(df, df_store) df_test = feat_matrix.loc[feat_matrix['Type'] == 'test'].copy() results = [] for top in TRIALS: predict = forecast(df_test, features_x, ENSEMBLE, top) save_result(df_test, predict) summit(sub_msg(top)) result = get_kaggle_score(sub_msg(top)) results.append((top, result)) print(results)
def main(): df, df_store = load_data(debug=False) feat_matrix, features_x, feature_y = extract_features(df, df_store) tune_random_forest(feat_matrix.loc[~(feat_matrix['Type'] == 'test')], features_x, feature_y)
def main(): df, df_store = load_data(debug=False) feat_matrix, features_x, feature_y = extract_features(df, df_store) train_ensemble(feat_matrix.loc[~(feat_matrix['Type'] == 'test')], features_x, feature_y)