sample users from follower bins

[bellecarrell]

• [x] get all unique users' json from 2018 sample

I would recommend finding follower bins, and sampling users before doing anything else

• [x] calculate follower bin percentiles for unique users

If you could sample 200 from each bin, and make sure that these users were currently active (query API for info) That will give us 1500-2000 users total There's another way to check if a user is still active, running through email to recover that now Like so: import requests requests.get('https://twitter.com/intent/user?user_id=3499399999') If the status code is 200, then the users should be active, else they have gone private or are deleted/banned

• [x] shuffle json strings + add first 200 for each range if active

• [x] make sure user is active

• [x] test/debug

[bellecarrell]

going to test/debug tomorrow

[bellecarrell]

sample users in /exp/acarrell/twitter_brand/blogger_2018/agg_17/

[bellecarrell]

forgot to add check that user is active -- will fix tomorrow

[bellecarrell]

updated sample users in /exp/acarrell/twitter_brand/blogger_2018/agg_17/

[bellecarrell]

forgot to save id files as well as files w/ full json. will fix today/tomorrow

[bellecarrell]

going to close + create new issue based on conversation yesterday