Getting error Check failed: (best_split_info.left_count) > (0) at /__w/1/s/python-package/compile/src/treelearner/serial_tree_learner.cpp,

[AnjaliSetiya]

Description

Reproducible example

Environment info

LightGBM version or commit hash:

Command(s) you used to install LightGBM

```I am getting the error on training data in LGBM. I am not able to find in teh documentation why this is happening. For earlier data this was not happening. Please let me know ho can this be resolved.

<!-- Put any additional environment information here -->


## Additional Comments
<!-- What else should we know? -->
****

[jameslamb]

Thanks for using LightGBM!

We're happy to help, but need you to provide more information.

Unfortunately with only an error message, it's not possible for us to help you without a very significant investment of effort and guesswork. For example, you haven't told us whether you are using Python, R, Java, C, etc.

Please provide the information that was asked for, like a reproducible example (code that demonstrates the problem), version of LightGBM you are using, how you installed it, etc.

[jameslamb]

I think you may also find https://docs.github.com/en/get-started/writing-on-github/getting-started-with-writing-and-formatting-on-github/basic-writing-and-formatting-syntax helpful for formatting text on GitHub.

[samFarrellDay]

@AnjaliSetiya I get this error when I have set min_sum_hessian_in_leaf == 0. To fix it, I set min_sum_hessian_in_leaf equal to some very small value, like 0.0000001. This might work in your case. I encounter it very rarely, and only because miceforest builds so many lightgbm models. @jameslamb there are several open issues referencing this error. #4739, #4946, #4286, #3679. This bug also looks like it's being tracked on the 4.0.0 release.

[jameslamb]

This bug

Thanks for linking those!

But to be clear, an error message is not the same as a "bug". Without more information from @AnjaliSetiya , it's not possible to determine whether the issue they're facing shares the same root cause as those others you've linked.

[AnjaliSetiya]

In my case I'm using the library in Python. I have dataset of around 10000 rows and 12000 columns. I'm trying to standardize the model. I did the hyperpatameter tuning earlier on a particular dataset. With those range of hyperpatameters that I have, I want to compare accuracy of a slightly different data, let's call it data2. For that purpose I was using the model with the same hyperpatameter range for data 2 is when I got this error. I didn't get the error on intial data. Min_data_in_hessian_leaf has been set to range 0 to 0.0001. Just to add I am using Bayesian optimisation for hyper-parameter tuning
Let me know if you have any solution.

[jameslamb]

Thanks for that! But that is still not enough information for us to reproduce this issue, or to suggest anything you can try to avoid it.

If you aren't able to provide the information I asked for in #5179 (comment), especially a reproducible example, then it isn't possible for us to tell whether whatever you're seeing is a different issue than others linked in #5179 (comment), and we'll close this issue.

[github-actions]

This issue has been automatically closed because it has been awaiting a response for too long. When you have time to to work with the maintainers to resolve this issue, please post a new comment and it will be re-opened. If the issue has been locked for editing by the time you return to it, please open a new issue and reference this one. Thank you for taking the time to improve LightGBM!

[AnjaliSetiya]

In my case I'm using the library in Python. I have dataset of around 10000 rows and 12000 columns. I'm trying to standardize the model. I did the hyperpatameter tuning earlier on a particular dataset. With those range of hyperpatameters that I have, I want to compare accuracy of a slightly different data, let's call it data2. For that purpose I was using the model with the same hyperpatameter range for data 2 is when I got this error. I didn't get the error on intial data. Min_data_in_hessian_leaf has been set to range 0 to 0.0001. Let me know if you have any solution.

…

On Tue, Apr 26, 2022, 18:13 Samuel Wilson ***@***.***> wrote: @AnjaliSetiya <https://github.com/AnjaliSetiya> I get this error when I have set min_sum_hessian_in_leaf == 0. To fix it, I set min_sum_hessian_in_leaf equal to some very small value, like 0.0000001. This might work in your case. @jameslamb <https://github.com/jameslamb> there are several open issues referencing this error. #4739 <#4739>, #4946 <#4946>, #4286 <#4286>, #3679 <#3679> — Reply to this email directly, view it on GitHub <#5179 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AP3RFJNPTY7NY73ACYWJNI3VG7QHLANCNFSM5UKWSPYA> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[jameslamb]

As we told you in #5179 (comment), that type of information is not sufficient for maintainers to investigate this report.

[github-actions]

This issue has been automatically locked since there has not been any recent activity since it was closed. To start a new related discussion, open a new issue at https://github.com/microsoft/LightGBM/issues including a reference to this.