qsd-group / dmsan Goto Github PK

Due to the lack of contextual parameter data

See these two figures:
https://github.com/QSD-Group/DMsan/blob/main/dmsan/bwaise/figures/improvements_heatmapA.png

https://github.com/QSD-Group/DMsan/blob/main/dmsan/bwaise/figures/improvements_heatmapC.png

Both of them are showing the winning chances, but have different scales, would be good to use the same scale.

I'm thinking as I type and also give you @haclohman an idea of what needs to be done next. I currently have managed to pull the different systems in a same script and made sure consistent parameter values are used in simulation, but I ran into this problem QSD-Group/EXPOsan#24 and need to take care of it first.

So next steps

• Resolve QSD-Group/EXPOsan#24, and hopefully we'll be able to get the scores for MCDA
• Update uncertainty_sensitivity.py for MCDA under uncertainty, I think we've got the uncertainty part figured out, but I stopped at sensitivity since I wanted to make sure that consistent parameter values are used and that led to the whole restructuring of the systems in EXPOsan, but now we should pick this up, I'll just add Spearman for now, I think from bwaise's experience KS didn't really provide much new info, we can add it later if we want to explore.
• If there's any analyses from bwaise you want to do, we can add those, but I'll want to develop generalizable functions instead of copying bwaise's codes and paste to another module for easier management and better consistency