I built the LightGBM binaries from https://github.com/microsoft/LightGBM. The build was successful, but the python API did not work for me. I am using the light gbm binary directly to run for time being.

I tested the binaries on LightGBM/examples/regression. The train and predict work fine using

lightgbm config=train.conf
lightgbm config=predict.conf

The outputs are _model.txt and _predict_result.txt

The problem I am facing is I cannot make out which in the input file regression.train is the target variable.

Also, I am looking for some documentation that explains how to run LightGBM without python API.

Thanks for the help.


1 Answer 1


The first column in the input file is considered to be the target variable. I found this by trial and error.


There is a parameter called label_column which can take an integer value from 0 to N, to specify which column is the target variable. (source: https://lightgbm.readthedocs.io/en/latest/Parameters.html)


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.