We are getting duplicates of random 5 char when we generate thousands of them at a time.
I just checked in the fix in frads. Please upgrade with pip install -U git+https://github.com/lbnl-eta/frads, you will need to overwrite the rfluxmtx file again.
It took 10min on my m1 Macbook pro to finish the run.
And I don’t think summing all the individual differential surface to differential surface view factors is correct, but I don’t know what would be the right way.