Skip to content
New issue

Have a question about this project? # for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “#”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? # to your account

add text-mined element embeddings to matminer via ElementProperty #343

Closed
computron opened this issue Dec 10, 2018 · 2 comments
Closed

add text-mined element embeddings to matminer via ElementProperty #343

computron opened this issue Dec 10, 2018 · 2 comments

Comments

@computron
Copy link
Contributor

It would be nice to have text-mined element embeddings in matminer. A whole separate study might compare these against MAGPIE-style descriptors.

The general strategy would be:

  • Add the raw embedding data to matminer.utils.data_files
  • Implement the embedding data in matminer.utils.data as a subclass of the AbstractData object
  • Implement the embedding featurize as a preset value in ElementProperty Featurizer in matminer.featurizers.composition. Once you have the embeddings as "AbstractData" above this step should be simple.

@ardunn can you work with Leigh on this?

@ardunn ardunn added the ugrads label Dec 11, 2018
@ardunn
Copy link
Contributor

ardunn commented Dec 16, 2018

Sure, what would you think about also adding a separate (but related) festurizer for the word embeddings of entire materials? leigh mentioned he could also give the embeddings for all materials, so this featurizer would just return the word embedding for a formula and would throw an error if the formula was not found/known

@ardunn
Copy link
Contributor

ardunn commented Dec 19, 2018

this is done

@ardunn ardunn closed this as completed Dec 19, 2018
# for free to join this conversation on GitHub. Already have an account? # to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants