-
Notifications
You must be signed in to change notification settings - Fork 15
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[ENH] Skeleton learning method #20
Conversation
Codecov Report
@@ Coverage Diff @@
## main #20 +/- ##
==========================================
- Coverage 76.83% 75.21% -1.63%
==========================================
Files 8 11 +3
Lines 367 585 +218
Branches 59 116 +57
==========================================
+ Hits 282 440 +158
- Misses 66 104 +38
- Partials 19 41 +22
📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more |
This now gives us a generic backbone for all the PC/FCI algo variants, where the first step is "learning a skeleton" from CI testing of the data. The |
# the tests get the correct independences when alpha is at 0.01 | ||
alpha = 0.01 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Hmm, 'tuning' the parameter in a way to make the test work is maybe not a good idea. There might be something wrong with the implementation, which is then not visible due to the tuned parameter. If this is purely due to the flakiness of the (statistical) tests, consider rerunning them them when they fail with newly generated data. If it fails, e.g., 3x in a row, then it is a good indication to double check the implementation (and this could be the independence test here). See for instance https://github.com/py-why/dowhy/blob/main/tests/gcm/independence_test/test_kernel.py#L12, where we use flaky for exactly these kind of flaky scenarios.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
IIRC I think I took the data from pcalg of the R package... so it's not actually generated on the fly.
I assumed it was fine since the tests passed other types of simulated data, but let me actually investigate.
I might just play around w/ simulating some new data instead from an SCM.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Oh wait I just checked again and 0.05 works. I think this might've been a different private project that I was working on. The alpha level works for 0.05 (default) and 0.01.
Signed-off-by: Adam Li <adam2392@gmail.com>
Signed-off-by: Adam Li <adam2392@gmail.com>
Signed-off-by: Adam Li <adam2392@gmail.com>
Signed-off-by: Adam Li <adam2392@gmail.com>
Signed-off-by: Adam Li <adam2392@gmail.com>
* Fix docs and certain ci builds for pygraphviz Signed-off-by: Adam Li <adam2392@gmail.com> * Add pygraphviz Signed-off-by: Adam Li <adam2392@gmail.com> * Add pygraphviz Signed-off-by: Adam Li <adam2392@gmail.com> Signed-off-by: Adam Li <adam2392@gmail.com>
Signed-off-by: Adam Li <adam2392@gmail.com>
Signed-off-by: Adam Li <adam2392@gmail.com>
Signed-off-by: Adam Li <adam2392@gmail.com>
Signed-off-by: Adam Li <adam2392@gmail.com>
Signed-off-by: Adam Li <adam2392@gmail.com>
Signed-off-by: Adam Li <adam2392@gmail.com>
Signed-off-by: Adam Li <adam2392@gmail.com>
Signed-off-by: Adam Li <adam2392@gmail.com>
Bumps [abatilo/actions-poetry](https://github.com/abatilo/actions-poetry) from 2.0.0 to 2.1.5. - [Release notes](https://github.com/abatilo/actions-poetry/releases) - [Changelog](https://github.com/abatilo/actions-poetry/blob/master/.releaserc) - [Commits](abatilo/actions-poetry@v2.0.0...v2.1.5) --- updated-dependencies: - dependency-name: abatilo/actions-poetry dependency-type: direct:production update-type: version-update:semver-minor ... Signed-off-by: dependabot[bot] <support@github.com> Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Signed-off-by: Adam Li <adam2392@gmail.com>
… tests (py-why#25) - Improve type checking for CI tests - add fingerprint for circleCI Signed-off-by: Adam Li <adam2392@gmail.com>
Signed-off-by: Adam Li <adam2392@gmail.com>
Signed-off-by: Adam Li <adam2392@gmail.com>
Signed-off-by: Adam Li <adam2392@gmail.com>
Signed-off-by: Adam Li <adam2392@gmail.com>
Signed-off-by: Adam Li <adam2392@gmail.com>
Follow on to: #19
And part of #18
Changes proposed in this pull request:
Before submitting
section of the
CONTRIBUTING
docs.Writing docstrings section of the
CONTRIBUTING
docs.After submitting