Hi,
I have just started looking at the Matminer. I am interested in calculating the Miedema features which is
matminer.featurizers.composition.``Miedema
(struct_types=‘inter’ , ss_types=‘min’ , data_source=‘Miedema’ )
I am trying to incorporate the Miedema into the following tutorial:
https://nbviewer.jupyter.org/github/hackingmaterials/matminer_examples/blob/master/matminer_examples/machine_learning-nb/bulk_modulus.ipynb
Used something like this after importing Miedema
md_feat = Midema
df = md_feat.featurize_dataframe(df,“composition”)
It does not work. According to the documentation, it reads a file “Miedema.csv”. Do I have to provide that file or it comes with the code.
Any help will be very much appreciated.
Thanks
Debasis
ardunn
April 4, 2019, 10:51pm
2
Hi Debasis,
Answers to a few questions might help us debug this:
Did you pip install matminer
?
1b. If so, what version of matminer do you have (e.g., pip list | grep matminer)
If you go to your matminer/utils/data_files/ directory, is there a Miedema.csv file?
Thanks,
Alex
···
On Thursday, April 4, 2019 at 3:28:20 PM UTC-7, [email protected] wrote:
Hi,
I have just started looking at the Matminer. I am interested in calculating the Miedema features which is
matminer.featurizers.composition.``Miedema
(struct_types=‘inter’ , ss_types=‘min’ , data_source=‘Miedema’ )
I am trying to incorporate the Miedema into the following tutorial:
https://nbviewer.jupyter.org/github/hackingmaterials/matminer_examples/blob/master/matminer_examples/machine_learning-nb/bulk_modulus.ipynb
Used something like this after importing Miedema
md_feat = Midema
df = md_feat.featurize_dataframe(df,“composition”)
It does not work. According to the documentation, it reads a file “Miedema.csv”. Do I have to provide that file or it comes with the code.
Any help will be very much appreciated.
Thanks
Debasis
Hi Debasis,
Another quick question, did you add parentheses after Miedema like this: md_feat = Miedema(), if you would like to use the default struct_types or ss_types?
Yeah, as Alex said, the “Miedema.csv” is already placed in the matminer/utils/data_files/ directory. I guess I can make a short PR to state it more clearly in the docs (and fix some other minor doc issues).
Thanks,
Qi
···
On Thursday, April 4, 2019 at 3:28:20 PM UTC-7, [email protected] wrote:
Hi,
I have just started looking at the Matminer. I am interested in calculating the Miedema features which is
matminer.featurizers.composition.``Miedema
(struct_types=‘inter’ , ss_types=‘min’ , data_source=‘Miedema’ )
I am trying to incorporate the Miedema into the following tutorial:
https://nbviewer.jupyter.org/github/hackingmaterials/matminer_examples/blob/master/matminer_examples/machine_learning-nb/bulk_modulus.ipynb
Used something like this after importing Miedema
md_feat = Midema
df = md_feat.featurize_dataframe(df,“composition”)
It does not work. According to the documentation, it reads a file “Miedema.csv”. Do I have to provide that file or it comes with the code.
Any help will be very much appreciated.
Thanks
Debasis
Hi Alex and Qi,
It did work by setting
md_feat = Midema()
I missed that parentheses.
I have a follow up question in the example link. It said the following:
"
Let’s try the ElectronegativityDiff
featurizer which requires knowing the oxidation state of the various elements in the Composition. This information is not there currently, but one can use the conversions
package to try to guess oxidation states and then apply the ElectronegativityDiff
featurizer to this column."
However it never really calculated the ElectronegativityDiff.
I used
`el_feat = ```ElectronegativityDiff()``
df = el_feat.featurize_dataframe(df,"composition_ox")
It came out with long list of errors. Mostly likely I am not using the right arguments.
Any help will be appreciated.
Debasis
···
On Thursday, April 4, 2019 at 6:28:20 PM UTC-4, [email protected] wrote:
Hi,
I have just started looking at the Matminer. I am interested in calculating the Miedema features which is
matminer.featurizers.composition.``Miedema
(struct_types=‘inter’ , ss_types=‘min’ , data_source=‘Miedema’ )
I am trying to incorporate the Miedema into the following tutorial:
https://nbviewer.jupyter.org/github/hackingmaterials/matminer_examples/blob/master/matminer_examples/machine_learning-nb/bulk_modulus.ipynb
Used something like this after importing Miedema
md_feat = Midema
df = md_feat.featurize_dataframe(df,“composition”)
It does not work. According to the documentation, it reads a file “Miedema.csv”. Do I have to provide that file or it comes with the code.
Any help will be very much appreciated.
Thanks
Debasis