Potential issue with mismatch version of pandas and matplotlib

[shchen-idmod]

In staging ssmt package, we have pandas 2.0.2 with matplotlib 3.2.2. these 2 combination does not work anymore. It will cause this error :
Reduce for Analyzer NodeCSVAnalyzer failed
Multi-dimensional indexing (e.g. obj[:, None]) is no longer supported. Convert to a numpy array before indexing instead.

There is very similar with this bug on internet. mwaskom/seaborn#3312

You can repro this issue with local AnalyzerManager test under https://github.com/InstituteforDiseaseModeling/idmtools/blob/master/idmtools_platform_comps/tests/inputs/run_multiple_analyzers_single_script.py (with latest pandas 2.0.2 and matplotlib 3.2.2)

Current production ssmt image was built with pandas 1.5.3 and matplotlib 3.2.2 so there is no problem. But next release of idmtools, it will pick up the latest pandas 2.0.2 (since we define in core with pandas>=1.1.4) so we need to lift restriction on requirements.txt under ssm-image for matplotlib:

idmtools/idmtools_platform_comps/ssmt_image/requirements.txt

Line 8 in 025f86c

matplotlib~=3.2.2

with latest version of matplotlib (matplotlib 3.7.1 seems working)

[shchen-idmod]

Fixed in 1.7.7