SepiaFullPrediction fails with no-discrepancy model

lanl / SEPIA

Simulation-Enabled Prediction, Inference, and Analysis: physics-informed statistical learning.

Other

29 stars 6 forks source link

SepiaFullPrediction fails with no-discrepancy model #17

Closed natalieklein229 closed 4 years ago

natalieklein229 commented 4 years ago

Error trace:

  File "sepia/SepiaPredict.py", line 211, in __init__
    uvPred(self)
  File "sepia/SepiaPredict.py", line 377, in uvPred
    betaV = samples['betaV'][ii, :]
KeyError: 'betaV'

In theory, FullPred should still work but should be able to check if no-discrepancy and handle that case.

jgattiker commented 4 years ago

The no-discrepancy scalar-only model is a new addition (not in GPMSA). Method not validated against gpmsa, or with real data.