Generating Synthetic History

[Şevval FINDIK]

Hello,

I want to generate synthetic history and I have data for 4 different years. I get the error below during train process. 

ValueError: autodetected range of [nan, nan] is not finite

I think the RuntimeWarnings I get during the fft process is the cause of the ValueError in the train.xml. However, I can't solve this. I thought about adding the lines below to the FastFourierTransform.py because I got '/usr/local/lib/python3.8/dist-packages/ravenframework/Models/PostProcessors/FastFourierTransform.py:135: RuntimeWarning: divide by zero encountered in true_divide rlz[target+'_fft_period'] = 1.0/freq' but system didn't let me. 

zero_indices = np.where(freq == 0) 

 freq[zero_indices] = np.finfo(float).eps

I appreciate if you can provide some insights.

Regards,

Şevval

[Gabriel J. Soto Gonzalez]

Hi Şevval,

 

Would you be able to share the input XML and other input files you are using? It would help in reproducing the error.

 

Thanks,

Gabriel

--
You received this message because you are subscribed to the Google Groups "INL RAVEN Users Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to inl-raven-use...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/inl-raven-users/159b3929-02c0-4ee2-a62c-b02d265f55aan%40googlegroups.com.

[Şevval FINDIK]

I have shared the files in the attachment.

Thanks,

Şevval

15 Kasım 2023 Çarşamba tarihinde saat 21:52:41 UTC+8 itibarıyla Gabriel J. Soto Gonzalez şunları yazdı:

[Gabriel J. Soto Gonzalez]

Hi Şevval,

Thanks for providing the files.

I tried running the train.xml file and it ran OK for me without errors (I got an arma.pk file, synthetic.csv, and the accompanying meta XML files).

I ran the fft.xml and it also finished running and returned what seemed to be the proper csv files, though I did see the “divide by zero” warnings in the terminal when I ran it. It doesn’t seem like the issues are linked.

Would you be able to provide the full stack error message (i.e., everything above the “ValueError: autodetected range of [nan, nan] is not finite” message)?

 

Updating to the newest RAVEN version might also help mitigate some problems (there are instructions in the RAVEN wiki here: https://github.com/idaholab/raven/wiki/Updating-and-testing-RAVEN).

Best,

To view this discussion on the web visit https://groups.google.com/d/msgid/inl-raven-users/2b461a13-2b4c-4ca0-84cf-c19262286ce5n%40googlegroups.com.

[Şevval FINDIK]

I shared them what I have after running fft.xml and train.xml in the attachment. 

Kind regards,

Şevval

17 Kasım 2023 Cuma tarihinde saat 04:30:05 UTC+8 itibarıyla Gabriel J. Soto Gonzalez şunları yazdı: