Nimish Purohit
Nimish Purohit
@Spinachboul There are numerous lines in the code that are not covered by the tests, so adding more tests would definitely help. Whenever you get to adding those tests, codecov...
I'll give it a go from my side, reading up a little on the algorithm.
The PCMCI algorithm looks alright to me, I'll see if I can find anything else.
@Spinachboul Somewhere along the line, a scalar value is being compared with an array. Looks like your call to the CI test for `pearsonr` is comparing `p_values` with different dimensions.
[This line](https://github.com/pgmpy/pgmpy/blob/dev/pgmpy/estimators/CITests.py#L506) in CI tests is causing the error. The `scipy.stats.pearsonr` function seems to be returning an array of p_values. You could use an example dataset to see how it...
Yeah, that's correct. I'll assign the issue to you.
I'll take a look. You can't merge stuff by yourself without review.
@narutonamikaze Your PR is for your own fork of `pgmpy`. You need to create a PR from your fork's feature branch (where you made the changes) to the dev branch...
No issue 🙂
@ankurankan I think this code should show the issue, you pointed it out some time ago - ```python from pgmpy.metrics import SHD from pgmpy.estimators import PC, ExpertKnowledge from pgmpy.utils import...