Current Wish List

To-do list 2024-1-25, to finish by 2024-4-31:

Will delete items when done.

1. Add Markov Checker function to Causal-Command (Kevin)
   • In progress. The Tetrad side of this project works, I think

All requested items--If I'm leaving something out, let me know...

1. Requested by end-users (top priority):
   1. Add method Causal Command to allow Markov Checker to be run on <graph, data> pairs
2. Requested by Peter:
   1. Add adjustment set gadget
3. Requested by Bryan @Bryan, do you still want to do these things? In Tetrad or in Python?
   1. IGES
   2. Plates interface
   3. Operation pipelines in the interface
4. Requested by Clark:
   1. Unmixing algorithm
      1. Madelyn sent the code
      2. Look at Strobl
      3. Also, Xinghua–uses UMAP to demix the data into cell types
   2. Add knowledge of unobserved confounding.
   3. Allow/disallow bidirected edges between tiers
   4. Biwei’s general score
   5. Two Step
      1. We may still not have the necessary linear algebra functions to do this in Java. This is Matlab code.
   6. PCP
      1. I got some downvotes for this. Let me ask Peter.
   7. DM
   8. GIN
      1. This is already in causal-learn; do we need it in Tetrad as well?
   9. Conditional Gaussian PM, IM
   10. Put algorithm comparison into the interface
       1. This may be something I could do now (I couldn't easily before)
   11. Bayes time series parametrization
   12. IM for time series with latent variables
   13. Usage statistics
5. Requested by Joe
   1. New tests/scores for “more general” cases
   2. Pip-install for py-tetrad
      1. Should do
   3. Input a specific list of conditional independence facts closed under selected graphoid axioms
   4. Update the data reader documentation and causal command documentation.
   5. Move the Tetrad manual to ReadTheDocs
   6. Apple signing of jars
      1. I am Not sure about this anymore; no one complains currently... need some votes for it