Julia is great for statistics, especially if you need to implement custom methods, and scientific ML, e.g. inferring ODEs from data. But Python, which I dislik

Julia is great for statistics, especially if you need to implement custom methods, and scientific ML, e.g. inferring ODEs from data. But Python, which I dislike as a language, has unbeatable differentiable ML libraries (TF, Torch, JAX). Julia's are quite impressive given the few man-hours they have received, but will give you a lot of trouble if you try to do non-trivial things.