Fix Old Retweets (Complete)

[remyknox]

Started

Current is_retweet column has ~72k NAs. Need to look into is_retweet.R to see why this might be.

Old retweets have "RT @XXXX:" at the beginning of retweet and is truncatated at ~120 characters in and ending with "...". Look into changing all retweets to have the same body of text at the original tweet.

[remyknox]

Current WIP script can replace the body of each retweet with the original tweet text. Still looking into (struggling) why there are so many NAs in the is_retweet column.

[remyknox]

Issue with ~72k NAs in is_retweet column comes from line 61 and 62 of is_retweet.R. Swapping these lines fixed the issue.

Because RTs that start with "RT @xxxx:" go up to 2018-06-30 UTC, it is best if the RTs are flagged by code used to replaced the RT text. This will be documented in fix_old_retweets.R. Running this code on the master data frame warrants creating a new version of the master df (as suggested by @brunj7).

[remyknox]

• Replace old retweets––started with "RT @xxxx:" and ended with "..." because they were truncated by Twitter––with the the original tweet using fix_old_retweets.R
• [x] need to run fix_old_retweets.R on master data frame

[remyknox]

@brunj7 requesting review of fix_old_retweets.R to confirm this wants to be run on entire data frame.

[remyknox]

Dates where retweets that started with "RT @xxxx:" and ended with "..."

[remyknox]

v3 Data Frames Created

Version 3 data frames have been created using fix_old_retweets.R. Both RT and noRT data frames now have the following modifications:

• [x] archive data (purchased data) no longer has NAs in "is_retweet"
• [x] in "text" any retweets with "RT @xxxx:" have been edited so that they are identical to the original tweet

  Complete

[brunj7]

great job ! Let us talk about the best strategy moving forward.