Small fry for ADM explained article

[operdeck]

pdstools version checks

• [X] I have checked that this issue has not already been reported.

• [X] I have confirmed this bug exists on the latest version of pdstools.

Issue description

looking at https://pegasystems.github.io/pega-datascientist-tools/Python/articles/ADMExplained.html and comparing to https://pegasystems.github.io/pega-datascientist-tools/R/articles/adm-explained.html

• auto-hide the not-so-interesting-bits of the code blocks (like the plotting plumbing)
• can we beautify the tables, highlighting the columns/rows of importance
• in the final classifier the resulting bin is hard coded to 6
• the adjusted propensity in the table is wrong, is all the same for all bins

The order of the fields is different but the python version is preferable (sorted)

Reproducible example

See python doc

Expected behavior

At par with R version

Installed versions

Replace this line with the output of pdstools.show_versions(), leave the backticks in place

[yusufuyanik1]

Updated with the points you mentioned.

• the adjusted propensity in the table is wrong, is all the same for all bins: You can't see this point from the article but it is also fixed. You can check the source actual notebook with code: https://github.com/pegasystems/pega-datascientist-tools/blob/master/examples/articles/ADMExplained.ipynb
• Additional point: Urls are not hidden in the article, will be updated with the next PR

[operdeck]

Cool. I can't see it in the article because that only gets republished with a new version?

Op di 23 mei 2023 om 18:10 schreef yusufuyanik1 @.***>:

Updated https://pegasystems.github.io/pega-datascientist-tools/Python/articles/ADMExplained.html with the points you mentioned.

• the adjusted propensity in the table is wrong, is all the same for all bins: You can't see this point from the article but it is also fixed. You can check the source actual notebook with code: https://github.com/pegasystems/pega-datascientist-tools/blob/master/examples/articles/ADMExplained.ipynb
• Additional point: Urls are not hidden in the article, will be updated with the next PR

— Reply to this email directly, view it on GitHub https://github.com/pegasystems/pega-datascientist-tools/issues/95#issuecomment-1559748913, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABEVTKGZLAENECNPSKLSD7DXHTONTANCNFSM6AAAAAAYKGW6WA . You are receiving this because you authored the thread.Message ID: @.***>

[StijnKas]

The article should be fully in sync with whatever's there in the repo, it's not tied to any versions. Let's check tomorrow.

[operdeck]

R version changed with better formulae and selected a different predictor, one with AUC < the model itself

[StijnKas]

@operdeck @yusufuyanik1 just a check - have we fixed this one?

[yusufuyanik1]

@StijnKas Yes, we aligned them