arpanatnear/hcc_v0

HCC : high confidence cohort (Attribution Panel)

HCC_v0

HCC : high confidence cohort (Attribution Panel)

To come up with a cohort with high ping density for a selected country and duration

input data:

Currently using (daily) staypoint output data

HCC_vo_build_master_db.py : Based on the given inputs, this will fetch the staypoints data and write the summarised data.

steps

• Read the one day s.p. data for xmode,rtb and cubiq
• Create time index for pings with midnight start as the refrence
• ifa + index level aggregataion : pings_index, gh7s_index (unique)
• ifa : count_index (present) in that day , pings_day
• cnt_index vs cnt_ifa disn for that day
• Write daily_counts to s3 : Master DB for HCC tunning
• loop for any given date and number of days

HCC_vo_aggn_from_master_db.py : this will fetch the above master data, apply the condition inputs and write the resulatant data to s3.