import numpy as np
import pandas as pd
from sklearn import linear_model
import sys

#TODO: check for valid sys.argv (must be 2 files)

with open(sys.argv[1], 'r') as train_file:
    train_df = pd.read_csv(train_file, sep='\t')

rows_count = len(train_df.index)
features_cols = ['x' + str(i).zfill(3) for i in range(30)]
X = np.array(train_df[features_cols]) #features
Y = np.array(train_df['y'])
clf = linear_model.SGDClassifier(loss='log', n_iter=(10**6)/rows_count)
clf.fit(X, Y)

with open(sys.argv[2], 'r') as test_file:
    test_df = pd.read_csv(test_file, sep='\t')

test_X = test_df[features_cols]
test_Y = np.array(test_df['y'])
predict_Y = clf.predict(test_X)

print('predicted:', predict_Y)
print('real:', test_Y)

error = np.mean(predict_Y != test_Y)
print('error:', error)