How do I extract the atomic orbitals using AtomicOrbitals?
I make the following script and I get an error
from matminer.featurizers.composition import AtomicOrbitals
ao_feat = AtomicOrbitals ()
ao_feat.featurize_dataframe (df, “composition_oxid”)
Hey Luis,
Please post your code and the error you get. Also include the system, python version, and matminer version you are using.
I am able to successfully use AtomicOrbitals:
In:
from matminer.datasets import load_dataset
from matminer.featurizers.composition import AtomicOrbitals
from matminer.featurizers.conversions import StrToComposition, CompositionToOxidComposition
ao_feat = AtomicOrbitals()
stc = StrToComposition()
coc = CompositionToOxidComposition()
df = load_dataset("matbench_expt_gap")
df = df.rename(columns={"composition": "strcomposition"})
df = stc.featurize_dataframe(df, "strcomposition")
df = coc.featurize_dataframe(df, "composition", ignore_errors=True)
df = ao_feat.featurize_dataframe(df, "composition_oxid", ignore_errors=True, return_errors=True)
print(df[ao_feat.feature_labels()])
Out:
HOMO_character HOMO_element ... LUMO_energy gap_AO
0 p S ... -0.261676 0.000000
1 d W ... -0.220603 0.000000
2 NaN NaN ... NaN NaN
3 NaN NaN ... NaN NaN
4 s Ag ... -0.157407 0.000000
... ... ... ... ...
4599 p N ... -0.182464 0.083833
4600 s Zr ... -0.150673 0.011718
4601 d Ti ... -0.170010 0.000000
4602 d Ti ... -0.170010 0.000000
4603 d W ... -0.220603 0.000000
Note that some of them failed, mostly because they didn’t have integer subscripts in the composition. You can get around that pretty easily but I didn’t here. I would bet your problem is exactly that, the compositions didn’t have integer subscripts.
Thanks,
Alex
1 Like