def predict(self): ''' Output a predicted price, and compare to the initial listed of user-guessed price ''' testing_df = create_testing_df(self.df) y = int(testing_df.pop('price').values) X = testing_df.values prediction = int(self.rfr.predict(X)[0]) predict_statement = 'Your model-recommended price is: $%s' %prediction compare_statement = 'Your initial stated price was: $%s' %y return predict_statement, compare_statement
gs = GridSearchCV(Ridge(), pars, cv=cv) gs.fit(X, y) ridge = gs.best_estimator_ print gs.best_params_ print gs.best_score_ cPickle.dump(ridge, open('models/ridge.pkl', 'wb')) pars = {'max_depth': [5, 8, 10, 20, 50, 100], 'min_samples_split': [2, 3, 5, 10, 20]} gs = GridSearchCV(RFR(n_estimators=10), pars, cv=cv) gs.fit(X, y) rfr = gs.best_estimator_ print gs.best_params_ print gs.best_score_ cPickle.dump(rfr, open('models/rfr.pkl', 'wb')) return ridge, rfr if __name__ == '__main__': df = pd.read_csv('data/complete_df.csv', header=False) testing_df = create_testing_df(df) y = testing_df.pop('price').values X = testing_df.values ridge, rfr = grid_search(X, y) grid_search(X, y)