Rogério Gouvêa
Rogério Gouvêa
Implemented a clipping method after scaling the data. Some features may only have reasonable values for a subset that wasn't included in the training set and when applying the scaler...
In std calculation to get the uncertainty, when the targets are high values, the float16 which was used as default in my calculations, blows up. Included a cast to np.float32...
Introduces a safe_join function to allow for more customized edits of the featurizers while preserving the rule for duplicated names with sufix and prefix.
Regarding this code in featurizers.py: ``` def featurize(self, df: pd.DataFrame) -> pd.DataFrame: """Run all of the preset featurizers on the input dataframe. Arguments: df: the input dataframe with a `"structure"`...
Matminer2024FastFeaturizer is not in the dictionary in FEATURIZER_PRESETS. More importantly, when using for composition only it reverts to CompositionOnlyMatminer2023Featurizer, however the structure_featurizers are not empty, it then includes erroneously GlobalSymmetryFeatures|is_centrosymmetric...