purge files with secrets

[maxheld83]

same as #8

BFG: https://rtyley.github.io/bfg-repo-cleaner/

[Ahobert]

I was not sure whether I will destroy something for your pull request, @maxheld83 so I did not delete anything for now.

Critical files are:

• exclude_from_analysis.csv
• exclude_from_inst_level_analysis.csv

We do not want to remove them entirely this time. They have two columns, one titled pk_kb_inst and a second one. We would want to remove the files from the repo, generate csv files that have just the second column and add them under the same name again in the data-raw folder.

@maxheld83 if you finish your pull request and still have some time and energy left, maybe you could take care of this? Otherwise, I will look into it tomorrow morning.

[maxheld83]

I think a simple way to go about this would be to add the remaining column as a simple R vector in, say exclude_from_analysis.R, and then to source() that file in place of the current readr::read_csv().
That way, we can cleanly kill all the old filenames, and also, one column doesn't really need a csv anyway.

But that will require some small changes in the remainder of the script, since the input is then a vector, and not a tibble.

[maxheld83]

closed in favor of #42