Closed basaks closed 5 years ago
Now we have not only r2 score in the transformed targets scale, but we also have all the other scores supported in uncoverml in the target transform scale.
For a model with a standard deviation output you will get all of these in the scores.json:
smse_transformed = 0.47923256677640025 lins_ccc = 0.6782576696084297 r2_score = 0.5174018081761803 smse = 0.4825981918238197 mll = 0.5554504342500338 expvar_transformed = 0.531611887457396 mll_transformed = 0.8322164916070733 expvar = 0.5234628450325468 r2_score_transformed = 0.5207674332235998 lins_ccc_transformed = 0.6820184726717139
When standard deviation is not available, the json will be missing the mll and mll_transformed output.
Now we have not only r2 score in the transformed targets scale, but we also have all the other scores supported in uncoverml in the target transform scale.
For a model with a standard deviation output you will get all of these in the scores.json:
When standard deviation is not available, the json will be missing the mll and mll_transformed output.