Price elasticity and small sample sizes

Hello and congratulations for this great package!

Currently I am working on a project which is related to price elasticity of demand. So, I have no X variable, and the only thing I would like to know is the price elasticity. If I have understood correctly, by taking the logs of Y(quantity) and T(price), the constant elasticity can be given by the CATE intercept by calling the const_marginal_effect. But I am not sure if my approach is correct, given also that my Y has a binomial distribution.

Another issue that I have, is that each time I run my model, I obtain highly or moderately different results, due to the small sample size (e.g. 800 rows). In order to tackle this issue, I have tried to use a repeated cross validation (sklearn) with the cv parameter of SparseLinearDML, but it produces an error. My first models are tree based. Can you propose a different approach, or give me any advice on how to tackle this issue?

Thanks a lot!

Read the SparseLinearDML documentation and reproduce the reported cv failure using the small-sample, tree-based setup described. Clarify whether the requested elasticity interpretation and cross-validation approach are supported; done means documenting a reproducible recommendation or a verified explanation.