Patrick Schaefer

FillBurn producing odd results on DEM

1

I'm working in a region with pretty extensive mining. The provincial DEM is quite out of date, so I tried replacing an area of the DEM with some LIDAR data...

[Feature Request] Linear Trend Line

4

Would it be possible to add an option for linear trend lines as well as the default?

I left a note here: #35, but thought it might warrant its own issue (if this package is still being developed). `scale_pos_weight` is only for binary classification, is there an...

Update explain.R

2

updated to work with lightgbm package version 3.3.5

[R-package] lightgbm::lgb.model.dt.tree() error caused by lgb.dump() error with large models

12

## Description When models or data sets reach a certain level of complexity, the `lgb.dump()` will cause an error in R: `Error: R character strings are limited to 2^31-1 bytes`....

lightgbm.unify() fails when model or data are too big

1

Due to R not being able to handle large character strings, with a resulting error: `Error: R character strings are limited to 2^31-1 bytes`. See https://github.com/microsoft/LightGBM/issues/6288#issuecomment-1983592815

added support for case_weights

14

I noticed there wasn't much progress on integrating case_weights into lightgbm through bonsai so I've taken a crack at adding it. This is in reference to [65](https://github.com/tidymodels/bonsai/issues/65).

Exposed n_startup_trials, folds, and multivariate from Optuna

1

Similar to [#73](https://github.com/StatMixedML/XGBoostLSS/pull/73) from XGBoostLSS, but for LightGBMLSS. I've taken a stab at writing the unit-test, but I'm still learning python, so please let me know if I missed anything.