search

erdogant / bnlearn

Python package for Causal Discovery by learning the graphical structure of Bayesian networks. Structure Learning, Parameter Learning, Inferences, Sampling methods.
https://erdogant.github.io/bnlearn
Other
482 stars 46 forks source link

Question regarding the Sachs dataset #74

Closed adam2392 closed 12 months ago

adam2392 commented 1 year ago

I am trying to find documentation regarding the Sachs dataset that is preprocessed and downloaded from url = "https://www.bnlearn.com/book-crc/code/sachs.interventional.txt.gz".

I was wondering if you could point out the documentation that tells us how it was preprocessed compared to the raw data?

I am reading https://www.bnlearn.com/research/sachs05/ and it seems that there are 11 experiments done (no observational data)? but then the link that downloads the preprocessed dataset only contains 6 unique values in the INT column denoting the interventions. Thanks!

Happy to help with a PR if its a matter of connecting some documentation/etc.

erdogant commented 1 year ago

Perhaps the paper (and supplemental material) itself will be the best source to find all the preprocessing steps. But it is behind a paywall (still after 2005 🤦🏻‍♂️) and can not open it.