Add ROC curve example code to docs

[kelly-sovacool]

By request of multiple Schloss Lab members.

It also would be good to show how to pick a good representative seed (i.e. something close to the median AUC / other performance metric) to then plot.

We might like to create a new vignette on visualizing results. Then we could move the plots already in vignette("parallel") and vignette("tuning") there (probably in a separate PR than for this issue).

[kelly-sovacool]

We'll riff on what @courtneyarmour has done here: https://github.com/SchlossLab/2021-08-09_ROCcurves
We may want to add some helper functions so users can make ROC curves easily from run_ml() output.

[kelly-sovacool]

Example code for plotting AUROC & AUPRC curves: https://github.com/SchlossLab/Barron_IBD-CDI_2022/blob/main/code/plotting-functions.R

[kelly-sovacool]

Incorporate and refactor functions from here: https://github.com/SchlossLab/mikropml-snakemake-workflow/blob/321110f5a03ae8108c4d0863e7e69ad26f4dc972/workflow/scripts/plot_roc_curves.R