Dan Handwerker

Another lightweight option would be to do region filling (either by default or as another option). This would mean that no voxel would be removed from the mask that is...

The very rough draft version of the flowchart & GraphViz code are attached. Just to copy key points from the Gitter conversation where this started: The attached version is just...

I'll second the idea of a table. The trouble I was having in putting everything in one flow chart is that the first part (what's above) was really about getting...

To answer @emdupre's question, I think it would greatly strengthen the code to get it to a state where a flow chart can automatically be generated based on what options...

One additional note on the current draft is that PCA dimensionality reduction is a single black box. Given there are a few options for this in consideration, I figured it...

FWIW, I think the current flow chart IS what would look better in the code. Much of the headache in putting this together is how the metrics are calculated in...

The updated flow chart plus table for the decision tree metrics and a new flow chart for the decision tree itself are attached. Note that there are comments in the...

Replies to some of @tsalo's comment. - `kappa ratio` and `rho elbow` both involve information not known before the decision tree, but I can't think of a way to include...

I definitely plan to have a flowchart that can be dynamically generated and output with each decision tree that is run. I've set up the decision tree function outputs to...

This is still in progress, with the goal of integrating flowchart creation with the modularized decision tree: #475 & #476