Visualize performance for subset featurespaces #43
Conversation
gwaybio
commented
Oct 19, 2023
•
gwaybio
commented
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
|
@jenna-tomkinson - the recent commits update the PR curves for the balanced model. Thanks! |
| @@ -139,12 +167,20 @@ f1_score_df <- readr::read_tsv( | |||
| "Phenotypic_Class" = "c", | |||
| "data_split" = "c", | |||
| "shuffled" = "c", | |||
| "feature_type" = "c" | |||
| "feature_type" = "c", | |||
| "balance_type" = "c" | |||
There was a problem hiding this comment.
Based on the plots, I am assuming that the balance type is no longer included, so could this be removed from the code since we have shifted to class-balanced models?
There was a problem hiding this comment.
On this specific command, I am loading a dataset using readr. IMO, readr does even a better job than pandas at specifying data type, AKA dtype. This line that you are highlighting is reading in the column named "balance_type" (and other columns) as a character dtype.
It is true that we only use balanced models in this visualization, but in order to do so, we need to make sure the column is loaded properly (specifically, as a character), so that we can filter later.
…henotypic_profiling_model into visualize-subset-featurespaces
|
Thanks for the review @jenna-tomkinson ! I will go ahead and merge now. |
