Crash when leaf nodes have no samples

[taltstidl]

Thanks for the great library. The visualizations are truly great. However, we're running into a crash when some of the leaf nodes do not have any samples in the X_train and y_train data. What seems to happen is the following:

• When calling render = viz_model.view() all necessary legend/node/leaf files are created. However, as some of the leaf nodes have no samples, the check in

  dtreeviz/dtreeviz/trees.py

  Line 1261 in a3d02a5

  if sum(counts) == 0:

  comes into play and no .svg file is generated for that leaf.
• Later when calling render.save('file.svg'), dot tries to render the SVG, but fails as the dot code returned by view still references the SVG files even though they were not created.

Two possible solutions to this:

1. dtreeviz/dtreeviz/trees.py

   Line 1240 in a3d02a5

   def _class_leaf_viz(node: ShadowDecTreeNode,

   Here, render an empty SVG or similar so it can later be rendered (this would be my preferred option).

2. dtreeviz/dtreeviz/trees.py

   Line 607 in a3d02a5

   leaves.append(class_leaf_node(node))

   Here, do not add a leaf if no file could be created (would likely need to add a suitable return value to _class_leaf_viz).

[tlapusan]

Hi @taltstidl ,

Thanks for raising this issue. There was some work related to it but it wasn't done completely. ex. #299

I will take a look on it in the next days, it seems to be an issue also when the split nodes don't have samples.

[Crefok]

workaround:
if you use the Model Training Data to create the dtreeviz.model you should be finde. If you want to draw a sample or Test Data you can get this Messages like

Warning: No such file or directory while opening .../leafXXX_{pid}.svg
Error: No or improper image file=".../leafXXX_{pid}.svg"
in label of node leafXXX