import csv
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.linear_model import LinearRegression
from stop_words import get_stop_words
import os
import pandas as pd

def read_train_file(inDirectory):
    colnames = ['start_date', 'end_date', 'title', 'sort_title', 'data']
    df_train = pd.read_csv(inDirectory, sep="\t", names=colnames)
    return df_train

def read_evaluate_file(inDirectory):
    colnames = ['data']
    df_evaluate = pd.read_csv(inDirectory, sep="\t", names=colnames, quoting=csv.QUOTE_NONE, error_bad_lines=False)
    return df_evaluate

def train_date_mean(df):
    date_mean = (df['start_date'] + df['end_date']) / 2
    return date_mean

def linearRegresionTrain():
    df = read_train_file('train/train.tsv')
    date_mean_df = train_date_mean(df)
    train_vectorized_corpus = vectorizer.fit_transform(df['data'])
    reg.fit(train_vectorized_corpus, date_mean_df)

def linearRegresionEvaluate(inDirectory, outDirectory):
    df_evaluate = read_evaluate_file(inDirectory)
    evaluate_vectorized_corpus = vectorizer.transform(df_evaluate['data'])
    evaluate = reg.predict(evaluate_vectorized_corpus)
    with open(outDirectory, 'w') as file:
        for e in evaluate:
            file.write("%f\n" % e)


vectorizer = TfidfVectorizer(stop_words=get_stop_words('polish'))
reg = LinearRegression()
linearRegresionTrain()
linearRegresionEvaluate('dev-0/in.tsv', 'dev-0/out.tsv')
linearRegresionEvaluate('dev-1/in.tsv', 'dev-1/out.tsv')
linearRegresionEvaluate('test-A/in.tsv', 'test-A/out.tsv')


os.system("./geval -t dev-0")
os.system("./geval -t dev-1")