That was in order to make smaller number of read queries (approximately O(logN)) instead of reading the files for each individual day which is O(N)