import pandas as pd
import h5sparse

NROWS = 10000
features = pd.read_csv('sorted_featureset.csv', nrows=NROWS)

meta = pd.read_hdf("meta.h5", start=0, stop=NROWS, key="data")
combined = pd.concat([meta[['user_id','session_id','step','is_val','is_test','clickout_id']], features[['user_id','session_id','step','is_val','is_test','clickout_id']].rename(lambda x: str(x) + '_r', axis=1)], axis=1)