Memory issues when N is increased too much

[JFernandoGR]

Dear EconML support team.

I am using Causal Forests for calculating CATEs, and the main issue I am facing is that confidence intervals calculated (CIs) are too large with my dataset. After reading QA information, the first solution is increasing the number of trees (nestimators) until the CATEs variance is controlled. The problem is that having ~100K obs, 2K estimators pose large problems of memory. Apart from trying with other hyperparameters, could you provide tips for dealing with it? (I have set up njobs equal to -1).