Hello,
I’m trying to run the example in publication ‘Heagy et al., 2017 Casing Example’ and
I know that it is a 3D example so takes time, I allowed the computer to run and after 1 day has stopped
with the error :
SystemError: gstrf was called with invalid arguments
Since is a System error, is it because my computer is not good enough to run this example?
I have tried in an AMD Ryzen 5 3400G 3.7GHz with 13.9 GB of RAM and in a Intel Pentium CPU G3220 3GHz with 8GB of RAM
I am not sure what it is, but it shouldn’t be your machine, I just quickly run it on my laptop. It does not require too much RAM/CPU.
@lheagy will probably know more.
However, it is a good point regarding RAM requirement of the examples. @lheagy, we should probably add the line sphinx_gallery_conf = {'show_memory': True} to the conf.py file, so it would show RAM usage.
I did run it on my core i3, 8 GB ram laptop as well, and it works fine. It doesn’t take too much time.
@Ezra Have you run with the reRun=True? or just downloaded the results? For me works fine when the reRun=False, but when I change for True give-me an error
SystemError: gstrf was called with invalid arguments
@jcbarreto Here is what has happened to my computer, when I run the code with ‘reRun=True’ (at both line 1438 and line 1481). It stopped working. I can’t even do any other stuff, so I had to stop it.
@Ezra, Ok so now we are talking about the same thing. For me, I need to allow the computer to run (without doing anything else) for 1 day, and in the end, I have that error. I don’t know what the error means…
@jcbarreto In my opinion, the inversion calculation is very big, which cannot be done by our machine.
This is a bit of a complicated example… If you are looking for casing examples, this might be a good place to start: https://github.com/simpeg-research/heagy-2018-em-casing
Thank you @lheagy for your attention, but I’m more interested in the sensitivity. Is there any example in TDEM (1D or 2D) like the MT_tutorial_Appendix_A_MT1D_Sensitivity? I was wondering why is necessary to multiply by a vector of ones to extract the Jacobian? When we do that, how we know which parameter corresponds to each row/line in the matrix, like how do you know that J_back_ey = J[0, nrx:].reshape(nx, ny, order=“F”) in the Heagy et al., 2017 Casing Example? Is the sensitivity calculated in the log space or in linear space? I would like to reproduce image 6.5 from your PhD thesis, are this notebook available?
