Has anyone tried using both in the same script? like using TMVA in parallel process for large dataset?
I’m just curious if it has been done before or it is even feasible to do so?
Hi,
We are in the process of parallelising TMVA for large data set. For example already the new DNN can be trained in parallel on the CPU and also GPU.
For other methods we are working on introducing parallelisation both using multi-threads and multi-process but we will not support parallelization using PROOF
Lorenzo