Pandapower_converter bug fix: Resolve bug with broadcasting

[furqan463]

Changes proposed in this PR include:

The changes proposed in PR #320 fixed known issues but created a bug in pgm_input_line function of pandapower_converter. in case of no data in column c0_nf_per_km, _get_attr_pp returns a single element array as it states broadcasting is responsibility of caller, but broadcasting was not handled, which creates error in assignment statemetns.

I've added another test to the /validation/converter/test_pandapower_converter_input.py, copied from pandapower tutorial power-grid-model_powerflow.ipynb, which fails on existing code.

[furqan463]

@mgovers need your help with this, apart from above bug, I noticed reliability rating problem SonarQube due floating point equality. I replaced it with np.isclose which will resolve this issue but now a unit test is failing which I'm not sure how to fix.

[furqan463]

unit test issue resolved and code improved. pre-commit done.

[mgovers]

Hi @furqan463 , thank you so much again for the contribution. @nitbharambe can you maybe pick this up? I am unfortunately not able to today or tomorrow.

[nitbharambe]

Thanks again @furqan463 !

[furqan463]

@mgovers @nitbharambe there is still a reliability issue on pypi.
I've checked that 2 No. lines causing this issue can be removed, the field is auto initialised with zero even when np.divide has some False entries in where.

[mgovers]

@mgovers @nitbharambe there is still a reliability issue on pypi. I've checked that 2 No. lines causing this issue can be removed, the field is auto initialised with zero even when np.divide has some False entries in where.

hi @furqan463, good find. can you please open a PR for that?