wiki-historian/mean_from_train/run.py

import pandas as pd
from tqdm import tqdm
import numpy as np

#r = pd.read_csv('../train/expected.tsv', header = None,sep='\t')

years = list()
with tqdm(open('../train/expected.tsv'), total=1_000_000) as f_in:
    for line in f_in:
        y_min, y_max = line.rstrip().split(',')
        years.append(float(y_min))
        years.append(float(y_max))

mean = np.mean(years)
mean = str(mean)
#mean = str(float(r.mean()))

print('mean: ' + mean)

with open('../dev-0/out.tsv','w') as f_out, open('../dev-0/in.tsv', 'r') as f_in:
    for line in f_in:
        f_out.write(mean + '\n')

with open('../test-A/out.tsv','w') as f_out, open('../test-A/in.tsv', 'r') as f_in:
    for line in f_in:
        f_out.write(mean + '\n')
        
        
# test 3