# Learning to Rank with Categorical variables

**URL:** <https://discourse.julialang.org/t/learning-to-rank-with-categorical-variables/78735>\
**Category:** General Usage\
**Created:** [March 30, 2022, 11:56am UTC](https://discourse.julialang.org/t/learning-to-rank-with-categorical-variables/78735 "2022-03-30T11:56:35Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![Billpete002](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/billpete002/32/35091_2.png) [@Billpete002](https://discourse.julialang.org/u/Billpete002)\
**Post date:** [March 30, 2022, 11:56am UTC](https://discourse.julialang.org/t/learning-to-rank-with-categorical-variables/78735/1 "2022-03-30T11:56:35Z")

</div>

Hi all,

I have a question about which package should I use for running against categorical and numerical values that will rank the results similar to Catboost with ranking as the objective. I know there is pycall.jl but I am hoping for a 100% pure Julia solution. I didn’t see any options for this in EvoTrees. I’ve used XGBoost.jl and I COULD one-hot encode some of the categorical variables, but there are some instances (like connections to airports) where the combinations are too numerous to make XGBoost feasible.

The data is quite large with hundreds of millions of records each day so speed is also a concern.

Thoughts or insights?

PS I did see a couple deprecated packages like JLBoost which might have have had this feature.