import pandas as pd
import numpy as np

dtypes = {'ad_id': np.float32, 'clicked': np.int8}
print("1")
train = pd.read_csv("../input/clicks_train.csv", usecols=['ad_id','clicked'], dtype=dtypes)

print("2")
ad_likelihood = train.groupby('ad_id').clicked.agg(['count','sum','mean']).reset_index()
M = train.clicked.mean()
del train

print("3")
ad_likelihood['likelihood'] = (ad_likelihood['sum'] + 12*M) / (12 + ad_likelihood['count'])

print("4")
test = pd.read_csv("../input/clicks_test.csv")
test = test.merge(ad_likelihood, how='left')
test.likelihood.fillna(M, inplace=True)

print("5")
test.sort_values(['display_id','likelihood'], inplace=True, ascending=False)
print("6")
subm = test.groupby('display_id').ad_id.apply(lambda x: " ".join(map(str,x))).reset_index()
print("7")
subm.to_csv("subm.csv", index=False)
print("8")