[TMVA][ROOT-10734] Veto RNN and CNN tutorials if pymva not enabled

[stwunsch]

No description provided.

[phsft-bot]

Starting build on ROOT-debian10-i386/cxx14, ROOT-performance-centos7-multicore/default, ROOT-fedora29/python3, ROOT-fedora30/cxx14, ROOT-fedora31/noimt, ROOT-ubuntu16/nortcxxmod, mac1015/cxx17, windows10/cxx14
How to customize builds

[lmoneta]

I think the tutorial should not use automatically pymva in case it is not available.
I would not disable them.
The change should be in not calling
TMVA::PyMethodBase::PyInitialize()

Unfortunately I don't think we have a pre-processor macro for pymva. So we need something not really nice but like this:

int iret = gSystem->Load("libPyMVA");
 if (iret >=0) gROOT->ProcessLine("TMVA::PyMethodBase::PyInitialize()");

[phsft-bot]

Build failed on ROOT-performance-centos7-multicore/default.
Running on olhswep22.cern.ch:/data/sftnight/workspace/root-pullrequests-build
See cdash.
See console output.

Errors:

• [2020-05-06T14:07:01.995Z] 646/2022 Test [TDF] minor fixes #370: tutorial-dataframe-df103_NanoAODHiggsAnalysis .....................................................***Failed Error regular expression found in output. Regex=[: error:] 4.48 sec

Failing tests:

• projectroot.runtutorials.tutorial_dataframe_df103_NanoAODHiggsAnalysis

[stwunsch]

Wouldn't it be easier to just remove the pymva part? I think an explicit library load in a tutorial is a little bit weird, from the users perspective. It does not convey the message of a clean tutorial for a stable feature.

[lmoneta]

HI,
I think is nice to show in a tutorial that we can run different methods and TMVA and Keras at the same time.
We could add a pre-processing macro for pymva, something like R__HAS_PYMVA
or even better fix the loading of the library using the ROOT plugin-manager.

I will see if I can provide a PR for one of these tomorrow

[oshadura]

@stwunsch @lmoneta is this still something we would like to be merged?

[lmoneta]

I think we can close this PR. It is not needed anymore

[oshadura]

@stwunsch I am sorry I closed PR but please reopen it if you feel it is something still needed :)