else: res = model.predict(X) return res if __name__ == "__main__": data = pd.read_csv('training.csv') test_data = pd.read_csv('testing.csv') X, test_X, y = preprocessing(data, test_data) # print(X.shape, test_X.shape) if Use_library: model = build_model() model = fit_model(model, X, y) res = predict(model, test_X) else: model = LogisticRegression(X, y, alpha=0.1, num_iters=50, regularized=True, normalization='l2') params = model.train(X, y, np.unique(y)) classifedLabels = [] for eachData in test_X: classifedLabels.append(model.classify(eachData, params)) res = np.array(classifedLabels) if save_res: np.save('result', res) npy2csv('result.npy', "result" + str(np.random.randint(0, 1000)) + ".csv")