maint: updating cicd and start_mapdl files

[clatapie]

As title says.

[ansys-reviewer-bot]

Thanks for opening a Pull Request. If you want to perform a review write a comment saying:

@ansys-reviewer-bot review

[codecov]

Codecov Report

Merging #223 (58a5f13) into main (793dc9c) will increase coverage by 7.02%.
The diff coverage is 100.00%.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #223      +/-   ##
==========================================
+ Coverage   85.61%   92.63%   +7.02%     
==========================================
  Files           2        2              
  Lines         584      584              
==========================================
+ Hits          500      541      +41     
+ Misses         84       43      -41     

[RobPasMue]

As I mentioned in ansys/pymapdl-examples#178 it is an issue inside the container

When running in PyMAPDL (if I am not mistaken) - you run things inside the container directly and you install these libraries inside it as well. However, here, you are connecting to it. So whenever you perform the install, it is not being installed for the container AFAICT.

[germa89]

As I mentioned in ansys/pymapdl-examples#178 it is an issue inside the container

When running in PyMAPDL (if I am not mistaken) - you run things inside the container directly and you install these libraries inside it as well. However, here, you are connecting to it. So whenever you perform the install, it is not being installed for the container AFAICT.

The container hasn't changed in ages (9 months)... why now it is asking for a new library to run? It does not make sense...

[germa89]

Opting for avoiding the issue all together....

Let's use the new docker images (do not have this issue). Hence we won't be testing for versions older than v23.1

[germa89]

Fixed, images are being pulled and launched properly. I would apply the same changes to pymapdl-examples
https://github.com/ansys/pyansys-math/actions/runs/6956135863?pr=223

There are some tests not passing properly. I leave it up to you @clatapie

[clatapie]

The unit test test_solve_eigs_km was failing because of a change in the MAPDL output when version > 23.2

Output for version 23.2:

*****  INDEX OF DATA SETS ON RESULTS FILE  *****\n\n   SET   TIME/FREQ    LOAD STEP   SUBSTEP  CUMULATIVE\n     1  1475.1             1         1         1                  \n     2  1475.1             1         2         2                  \n     3  2018.8             1         3         3                  \n     4  2018.8             1         4         4                  \n     5  2018.8             1         5         5                  \n     6  2024.8             1         6         6                  \n     7  2024.8             1         7         7                  \n     8  2024.8             1         8         8                  \n     9  2242.2             1         9         9                  \n    10  2274.8             1        10        10

Output for version 24.1:

*****  INDEX OF DATA SETS ON RESULTS FILE  *****\n\n     SET   TIME/FREQ    LOAD STEP   SUBSTEP  CUMULATIVE\n       1  1475.1333             1         1         1                  \n       2  1475.1333             1         2         2                  \n       3  2018.8374             1         3         3                  \n       4  2018.8374             1         4         4                  \n       5  2018.8374             1         5         5                  \n       6  2024.7868             1         6         6                  \n       7  2024.7868             1         7         7                  \n       8  2024.7868             1         8         8                  \n       9  2242.2153             1         9         9                  \n      10  2274.8300             1        10        10

The number of digits after the decimal point increased from 1 (for 23.2) to 4 ( for24.1). The number were therefore no longer caught by the re.findall call.

[RobPasMue]

Ready to merge?