from sklearn.feature_extraction.text import TfidfVectorizer
import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.naive_bayes import GaussianNB

nrows = 10000


def load_data(path):
    in_df = pd.read_csv(f'{path}/in.tsv.xz', sep='\t', nrows=nrows, header=None)
    try:
        exp_df = pd.read_csv(f'{path}/expected.tsv', sep='\t', nrows=nrows, header=None)
    except:
        exp_df = None
    return in_df, exp_df


def write_res(data, path):
    with open(path, 'w') as f:
        for line in data:
            f.write(f'{line}\n')
    print(f"Data written {path}/out.tsv")




def main():
    in_df, exp_df = load_data('train')
    in_df = in_df.drop(in_df.columns[1], axis=1)
    
    vectorizer = TfidfVectorizer(lowercase=True, stop_words=['english'])
    X = vectorizer.fit_transform(in_df.to_numpy().ravel())
    vectorizer.get_feature_names_out()

    # in_df = in_df.reset_index()

    tfidfVector = vectorizer.transform(in_df[0])
    gnb = GaussianNB()
    gnb.fit(tfidfVector.todense(), exp_df)

    paths = ['dev-0', 'test-A']

    for path in paths:
        x = load_data(path)
        vec = vectorizer.transform(x[0][0])
    
        result = gnb.predict(vec.todense())
        write_res(result, f'{path}/out.tsv')



if __name__ == '__main__':
    main()