James Lamb

@ericstj sorry for the delayed response. We don't have much .NET experience here. Are you interested in submitting a fix?

@shiyu1994 could you please help with this question?

Thanks for using LightGBM. > LightGBM version or commit hash:latest > Command(s) you used to install LightGBM：Follow the official website tutorial These are not acceptable answers. "latest" is not a...

> could you also add a Python tag? We don't actually have a `python` label but probably should. I've at least edited the PR title for now.

hmmmm @btrotta @guolinke if `interaction_constraints` is provided, should features that do not appear be excluded from training? At first I agreed with @mayer79 's assessment that they **should not be...

> Is it an option to add a snippet like this? I'd prefer that that be done outside of `_param_dict_to_str()`, similar to how `categorical_feature` is resolved: https://github.com/microsoft/LightGBM/blob/28536a0a5d0c46598fbc930a63cec72399a969e4/python-package/lightgbm/basic.py#L2110-L2130 I'd support it...

I'm generally supportive of everything you've proposed at the top of this issue, thanks for working on it and for adding that context about consistency with XGBoost and scikit-learn! Let's...

😂 ok thank you, I don't remember either. Agree then, let's move forward @mayer79

Thanks for using LightGBM. Are you able to share some **minimal** code showing precisely what you mean? I'm unsure how to interpret some of these statements like "a very large...

Ok, so to clarify: * you are not actually using Random Forest mode? (`boosting="rf"`) * you are using the built-in `binary` loss function (`objective="binary"`)?