20 lines
684 B
Python
20 lines
684 B
Python
|
import pandas as pd
|
||
|
|
|
||
|
|
train_in = pd.read_csv('../train/in.tsv', header = None, names = ['id', 'nazwa','typ_zbioru', 'rok', 'miesiąc'], sep='\t')
|
||
|
|
train_exp = pd.read_csv('../train/expected.tsv', header = None, names = ['opad'], sep='\t')
|
||
|
|
|
||
|
|
train = train_in.copy()
|
||
|
|
train['opad'] = train_exp
|
||
|
|
|
||
|
|
|
||
|
|
by_miesiac = train.groupby(['rok', 'miesiąc']).mean()
|
||
|
|
|
||
|
|
|
||
|
|
for dataset in ('dev-0', 'test-A'):
|
||
|
|
with open(f'../{dataset}/out.tsv','w') as f_out, open(f'../{dataset}/in.tsv', 'r') as f_in:
|
||
|
|
for line in f_in:
|
||
|
|
_, _, _, rok, miesiac = line.rstrip().split('\t')
|
||
|
|
srednia = by_miesiac.loc[int(rok), int(miesiac)]['opad']
|
||
|
|
f_out.write(str(srednia) + '\n')
|
||
|
|
|