mismatching array sizes when reading element_solid_history_variables

[jakobstanford]

🐛 Describe the bug
I am trying to read element_solid_history_variables from my *.ptf files, but when I'm calling D3plot() with state_array_filter or state_filter, I get get an error.

🔢 To Reproduce

1. take any .ptf file (d3plot files seem to work)
2. run the following script:

from lasso.dyna import D3plot, ArrayType, FilterType
res = D3plot('GM108.ptf', state_array_filter=[ArrayType.element_solid_history_variables])

3. get this error message:
   ValueError: could not broadcast input array from shape (522,5913,1,11) into shape (522,5913,1,5)

💘 Expected behavior
Function call works without throwing an error message.

🖥️ Setup

• lasso-python version: 2.0.2
• OS: Win11 x64, but also occurs on Linux.

ℹ️ Additional context
My current workaround is to call D3plot() without arguments. I can then read out the array as expected, but the issue is that this doesn't work with the rest of my infrastructure, where I can only read a set of timesteps at a time.

[codie3611]

What is a ptf file?

[jakobstanford]

Its similar to d3plot files. I actually thought it is exactly the same with a different file extension. Except that simply renaming my GM108.ptf to d3plot doesnt make the problem go away. So there must be some difference.

https://help.oasys-software.com/articles/#!d3plot-21-0/the-complete-state-ptf-file-also-rlf-and-d3eigv-files

[honzatomek]

I created a pull request for this issue (actually noticed this issue after creating the pull request as I came up with the same problem with my d3plot files). You can check the solution here:
https://github.com/honzatomek/lasso-python/tree/bugfix/element_solid_strain

The problem is in _read_states_solid function in d3plot.py file where the i_solid_vars already contains the element_solid_history_variables therefore it is wrong to increment it again. The solid true strain data (element_solid_plastic_strain_tensor, element_solid_thermal_strain_tensor and element_solid_strain) are contained in element_solid_history_variables, if present. When element_solid_strain is requested, it is at the end of solid element data, therefore also at the end of element_solid_history_variables. The element_solid_history_variables are composed of strain data in the following order: [element_solid_plastic_strain_tensor], [element_solid_thermal_strain_tensor], [element_solid_strain].

Excerpt from the docu: