The python package has plotting functions to visualize the trees grown. Is there a plan to add the same functionalities to the R package? Something like xgboost's xgb.plot.tree and xgb.plot.multi.trees using DiagrammeR.
@yanyachen any idea?
These go above and beyond what's current in xgb.plot.tree; would be great to have them in R for LightGBM http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=8017582
A function similar to xgb.plot.multi.trees method would be especially useful for biological applications like setting gates for cell sorting, where model interpretation is pivotal.
@Laurae2 if no one is working this, I can help with the feature request.
@randxie In addition to my link above, another possible starting point is to extend SHAP (which is already implemented in python-lightgbm): https://github.com/slundberg/shap/blob/master/notebooks/Census%20income%20classification%20with%20LightGBM.ipynb
@randxie You can base the charts on the following R package if you want a baseline to compare with xgboost: https://github.com/dmlc/xgboost/tree/master/R-package/R
https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=8017582 is also extremely interesting and go way beyond what LightGBM currently has as @zkurtz pointed out earlier..
Interactive link: http://shixialiu.com:8082/static/index.html

Closed in favor of being in #2302. We decided to keep all feature requests in one place.
Welcome to contribute this feature! Please re-open this issue (or post a comment if you are not a topic starter) if you are actively working on implementing this feature.
Adding this comment to this discussion: https://github.com/microsoft/LightGBM/pull/3188#issuecomment-677278222
hey @randxie are you still interested in trying this :stuck_out_tongue: