Checking git conflict#72
Open
daehyun99 wants to merge 42 commits into
Open
Conversation
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request introduces version 1.1.0, featuring a comprehensive documentation overhaul, a refactored PDAG class, and a reimplemented GES causal discovery algorithm. It also adds causal statistics to get_stats, supports fixed-edge counts in random graph generation, and updates instructions for AI agents. Review feedback focuses on optimizing the new GES implementation by removing redundant acyclicity checks, replacing inefficient path searches with reachability tests, and deferring expensive graph conversions to improve execution speed.
daehyun99
force-pushed
the
Refactor/2385/PDAG-1
branch
from
ad7b1a8 to
7cb464b
Compare
daehyun99
force-pushed
the
Refactor/2385/PDAG-1
branch
from
7cb464b to
b740f85
Compare
5 tasks
**The following checklist is mandatory** Your PR will be closed if you remove the checklist. Use of LLMs is **strictly forbidden** for any part of this checklist (including for improving language), and will result in a **ban** if we find any use of LLMs. ### Your checklist for this pull request - [ ] Have you followed all the steps from our [Contributing Guide](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md)? - [ ] Does the PR fully address the linked issue and is within its defined scope? If you are still working on the PR, mark it as draft. - [ ] Are all the GitHub Actions checks passing? If not, mark your PR as draft while you fix it. If you have used AI/LLMs for any assistance, please answer the following questions. Please refer [pgmpy#2622](pgmpy#2622) for an example of the level of detail we expect: - [ ] Have you reviewed our [AI usage policy](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md#ai-usage-policy)? - [ ] Are you able to fully explain your changes? We expect you to fully understand the algorithm and take full responsibility for any changes in this PR. - Please describe in **detail** how and what you used AI assistance for? Please outline your whole workflow with the AI tool. [Please replace this with your answer] - What steps have you taken to verify that the changes correctly address the issue? What edge cases have you considered? Other than running tests, what else have you verified? [Please replace this with your answer] - Have you used AI for generating tests? Can you compress them into a smaller number of tests without losing coverage? [Please replace this with your answer] ### Issue number(s) that this pull request fixes - Fixes # ### List of changes to the codebase in this pull request -
**The following checklist is mandatory** Your PR will be closed if you remove the checklist. Use of LLMs is **strictly forbidden** for any part of this checklist (including for improving language), and will result in a **ban** if we find any use of LLMs. ### Your checklist for this pull request - [ ] Have you followed all the steps from our [Contributing Guide](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md)? - [ ] Does the PR fully address the linked issue and is within its defined scope? If you are still working on the PR, mark it as draft. - [ ] Are all the GitHub Actions checks passing? If not, mark your PR as draft while you fix it. If you have used AI/LLMs for any assistance, please answer the following questions. Please refer [pgmpy#2622](pgmpy#2622) for an example of the level of detail we expect: - [ ] Have you reviewed our [AI usage policy](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md#ai-usage-policy)? - [ ] Are you able to fully explain your changes? We expect you to fully understand the algorithm and take full responsibility for any changes in this PR. - Please describe in **detail** how and what you used AI assistance for? Please outline your whole workflow with the AI tool. [Please replace this with your answer] - What steps have you taken to verify that the changes correctly address the issue? What edge cases have you considered? Other than running tests, what else have you verified? [Please replace this with your answer] - Have you used AI for generating tests? Can you compress them into a smaller number of tests without losing coverage? [Please replace this with your answer] ### Issue number(s) that this pull request fixes - Fixes # ### List of changes to the codebase in this pull request -
**The following checklist is mandatory** Your PR will be closed if you remove the checklist. Use of LLMs is **strictly forbidden** for any part of this checklist (including for improving language), and will result in a **ban** if we find any use of LLMs. ### Your checklist for this pull request - [ ] Have you followed all the steps from our [Contributing Guide](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md)? - [ ] Does the PR fully address the linked issue and is within its defined scope? If you are still working on the PR, mark it as draft. - [ ] Are all the GitHub Actions checks passing? If not, mark your PR as draft while you fix it. If you have used AI/LLMs for any assistance, please answer the following questions. Please refer [pgmpy#2622](pgmpy#2622) for an example of the level of detail we expect: - [ ] Have you reviewed our [AI usage policy](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md#ai-usage-policy)? - [ ] Are you able to fully explain your changes? We expect you to fully understand the algorithm and take full responsibility for any changes in this PR. - Please describe in **detail** how and what you used AI assistance for? Please outline your whole workflow with the AI tool. [Please replace this with your answer] - What steps have you taken to verify that the changes correctly address the issue? What edge cases have you considered? Other than running tests, what else have you verified? [Please replace this with your answer] - Have you used AI for generating tests? Can you compress them into a smaller number of tests without losing coverage? [Please replace this with your answer] ### Issue number(s) that this pull request fixes - Fixes # ### List of changes to the codebase in this pull request -
**The following checklist is mandatory** Your PR will be closed if you remove the checklist. Use of LLMs is **strictly forbidden** for any part of this checklist (including for improving language), and will result in a **ban** if we find any use of LLMs. ### Your checklist for this pull request - [ ] Have you followed all the steps from our [Contributing Guide](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md)? - [ ] Does the PR fully address the linked issue and is within its defined scope? If you are still working on the PR, mark it as draft. - [ ] Are all the GitHub Actions checks passing? If not, mark your PR as draft while you fix it. If you have used AI/LLMs for any assistance, please answer the following questions. Please refer [pgmpy#2622](pgmpy#2622) for an example of the level of detail we expect: - [ ] Have you reviewed our [AI usage policy](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md#ai-usage-policy)? - [ ] Are you able to fully explain your changes? We expect you to fully understand the algorithm and take full responsibility for any changes in this PR. - Please describe in **detail** how and what you used AI assistance for? Please outline your whole workflow with the AI tool. [Please replace this with your answer] - What steps have you taken to verify that the changes correctly address the issue? What edge cases have you considered? Other than running tests, what else have you verified? [Please replace this with your answer] - Have you used AI for generating tests? Can you compress them into a smaller number of tests without losing coverage? [Please replace this with your answer] ### Issue number(s) that this pull request fixes - Fixes # ### List of changes to the codebase in this pull request -
…mpy#3293) **The following checklist is mandatory** Your PR will be closed if you remove the checklist. Use of LLMs is **strictly forbidden** for any part of this checklist (including for improving language), and will result in a **ban** if we find any use of LLMs. ### Your checklist for this pull request - [x] Have you followed all the steps from our [Contributing Guide](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md)? - [x] Does the PR fully address the linked issue and is within its defined scope? If you are still working on the PR, mark it as draft. - [x] Are all the GitHub Actions checks passing? If not, mark your PR as draft while you fix it. If you have used AI/LLMs for any assistance, please answer the following questions. Please refer [pgmpy#2622](pgmpy#2622) for an example of the level of detail we expect: - [x] Have you reviewed our [AI usage policy](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md#ai-usage-policy)? - [x] Are you able to fully explain your changes? We expect you to fully understand the algorithm and take full responsibility for any changes in this PR. - Please describe in **detail** how and what you used AI assistance for? Please outline your whole workflow with the AI tool. Didn't use any AI for this PR - What steps have you taken to verify that the changes correctly address the issue? What edge cases have you considered? Other than running tests, what else have you verified? I ran the doctests multiple times locally - Have you used AI for generating tests? Can you compress them into a smaller number of tests without losing coverage? Not applicable ### Issue number(s) that this pull request fixes - Fixes pgmpy#3219 ### List of changes to the codebase in this pull request - Used load_model where were needed to load the model from blearn/asia to convert to asia.net and run the test - Updated doctests to use: `# doctest: +NORMALIZE_WHITESPACE` for formatting consistency --------- Co-authored-by: Ankur Ankan <ankurankan@gmail.com>
- Add missing pandas import in sample() docstring example - Fix TypeError by converting dtype string to torch.dtype using getattr - Update expected output to match current numpy/torch versions Fixes pgmpy#3154 **The following checklist is mandatory** Your PR will be closed if you remove the checklist. Use of LLMs is **strictly forbidden** for any part of this checklist (including for improving language), and will result in a **ban** if we find any use of LLMs. ### Your checklist for this pull request - [yes] Have you followed all the steps from our [Contributing Guide](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md)? - [yes] Does the PR fully address the linked issue and is within its defined scope? If you are still working on the PR, mark it as draft. - [ ] Are all the GitHub Actions checks passing? If not, mark your PR as draft while you fix it. If you have used AI/LLMs for any assistance, please answer the following questions. Please refer [pgmpy#2622](pgmpy#2622) for an example of the level of detail we expect: - [yes] Have you reviewed our [AI usage policy](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md#ai-usage-policy)? - [yes] Are you able to fully explain your changes? We expect you to fully understand the algorithm and take full responsibility for any changes in this PR. - Please describe in **detail** how and what you used AI assistance for? Please outline your whole workflow with the AI tool. I used Claude (an AI assistant) as a learning coach to help me understand the error and navigate the codebase. It helped me understand what getattr does and why torch.as_tensor needs a torch.dtype object instead of a string. I wrote and ran all the commands myself. - What steps have you taken to verify that the changes correctly address the issue? What edge cases have you considered? Other than running tests, what else have you verified? I ran python3 -m doctest pgmpy/factors/hybrid/FunctionalCPD.py -v and confirmed 13 tests pass. I also ran the sample function manually in a Python script to verify the output is correct. - Have you used AI for generating tests? Can you compress them into a smaller number of tests without losing coverage? No. This PR fixes doctest examples that already existed. I did not generate any new tests. ### Issue number(s) that this pull request fixes - Fixes pgmpy#3154 ### List of changes to the codebase in this pull request - pgmpy/factors/hybrid/FunctionalCPD.py --------- Co-authored-by: Ankur Ankan <ankurankan@gmail.com>
**The following checklist is mandatory** Your PR will be closed if you remove the checklist. Use of LLMs is **strictly forbidden** for any part of this checklist (including for improving language), and will result in a **ban** if we find any use of LLMs. ### Your checklist for this pull request - [ ] Have you followed all the steps from our [Contributing Guide](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md)? - [ ] Does the PR fully address the linked issue and is within its defined scope? If you are still working on the PR, mark it as draft. - [ ] Are all the GitHub Actions checks passing? If not, mark your PR as draft while you fix it. If you have used AI/LLMs for any assistance, please answer the following questions. Please refer [pgmpy#2622](pgmpy#2622) for an example of the level of detail we expect: - [ ] Have you reviewed our [AI usage policy](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md#ai-usage-policy)? - [ ] Are you able to fully explain your changes? We expect you to fully understand the algorithm and take full responsibility for any changes in this PR. - Please describe in **detail** how and what you used AI assistance for? Please outline your whole workflow with the AI tool. [Please replace this with your answer] - What steps have you taken to verify that the changes correctly address the issue? What edge cases have you considered? Other than running tests, what else have you verified? [Please replace this with your answer] - Have you used AI for generating tests? Can you compress them into a smaller number of tests without losing coverage? [Please replace this with your answer] ### Issue number(s) that this pull request fixes - Fixes # ### List of changes to the codebase in this pull request -
**The following checklist is mandatory** Your PR will be closed if you remove the checklist. Use of LLMs is **strictly forbidden** for any part of this checklist (including for improving language), and will result in a **ban** if we find any use of LLMs. ### Your checklist for this pull request - [ ] Have you followed all the steps from our [Contributing Guide](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md)? - [ ] Does the PR fully address the linked issue and is within its defined scope? If you are still working on the PR, mark it as draft. - [ ] Are all the GitHub Actions checks passing? If not, mark your PR as draft while you fix it. If you have used AI/LLMs for any assistance, please answer the following questions. Please refer [pgmpy#2622](pgmpy#2622) for an example of the level of detail we expect: - [ ] Have you reviewed our [AI usage policy](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md#ai-usage-policy)? - [ ] Are you able to fully explain your changes? We expect you to fully understand the algorithm and take full responsibility for any changes in this PR. - Please describe in **detail** how and what you used AI assistance for? Please outline your whole workflow with the AI tool. [Please replace this with your answer] - What steps have you taken to verify that the changes correctly address the issue? What edge cases have you considered? Other than running tests, what else have you verified? [Please replace this with your answer] - Have you used AI for generating tests? Can you compress them into a smaller number of tests without losing coverage? [Please replace this with your answer] ### Issue number(s) that this pull request fixes - Fixes # ### List of changes to the codebase in this pull request -
**The following checklist is mandatory** Your PR will be closed if you remove the checklist. Use of LLMs is **strictly forbidden** for any part of this checklist (including for improving language), and will result in a **ban** if we find any use of LLMs. ### Your checklist for this pull request - [ ] Have you followed all the steps from our [Contributing Guide](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md)? - [ ] Does the PR fully address the linked issue and is within its defined scope? If you are still working on the PR, mark it as draft. - [ ] Are all the GitHub Actions checks passing? If not, mark your PR as draft while you fix it. If you have used AI/LLMs for any assistance, please answer the following questions. Please refer [pgmpy#2622](pgmpy#2622) for an example of the level of detail we expect: - [ ] Have you reviewed our [AI usage policy](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md#ai-usage-policy)? - [ ] Are you able to fully explain your changes? We expect you to fully understand the algorithm and take full responsibility for any changes in this PR. - Please describe in **detail** how and what you used AI assistance for? Please outline your whole workflow with the AI tool. [Please replace this with your answer] - What steps have you taken to verify that the changes correctly address the issue? What edge cases have you considered? Other than running tests, what else have you verified? [Please replace this with your answer] - Have you used AI for generating tests? Can you compress them into a smaller number of tests without losing coverage? [Please replace this with your answer] ### Issue number(s) that this pull request fixes - Fixes # ### List of changes to the codebase in this pull request -
**The following checklist is mandatory** Your PR will be closed if you remove the checklist. Use of LLMs is **strictly forbidden** for any part of this checklist (including for improving language), and will result in a **ban** if we find any use of LLMs. ### Your checklist for this pull request - [ ] Have you followed all the steps from our [Contributing Guide](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md)? - [ ] Does the PR fully address the linked issue and is within its defined scope? If you are still working on the PR, mark it as draft. - [ ] Are all the GitHub Actions checks passing? If not, mark your PR as draft while you fix it. If you have used AI/LLMs for any assistance, please answer the following questions. Please refer [pgmpy#2622](pgmpy#2622) for an example of the level of detail we expect: - [ ] Have you reviewed our [AI usage policy](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md#ai-usage-policy)? - [ ] Are you able to fully explain your changes? We expect you to fully understand the algorithm and take full responsibility for any changes in this PR. - Please describe in **detail** how and what you used AI assistance for? Please outline your whole workflow with the AI tool. [Please replace this with your answer] - What steps have you taken to verify that the changes correctly address the issue? What edge cases have you considered? Other than running tests, what else have you verified? [Please replace this with your answer] - Have you used AI for generating tests? Can you compress them into a smaller number of tests without losing coverage? [Please replace this with your answer] ### Issue number(s) that this pull request fixes - Fixes # ### List of changes to the codebase in this pull request -
Auto-generated by the nightly CI workflow. Merging this will improve test split balancing. Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
**The following checklist is mandatory** Your PR will be closed if you remove the checklist. Use of LLMs is **strictly forbidden** for any part of this checklist (including for improving language), and will result in a **ban** if we find any use of LLMs. ### Your checklist for this pull request - [x] Have you followed all the steps from our [Contributing Guide](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md)? - [x] Does the PR fully address the linked issue and is within its defined scope? If you are still working on the PR, mark it as draft. - [x] Are all the GitHub Actions checks passing? If not, mark your PR as draft while you fix it. If you have used AI/LLMs for any assistance, please answer the following questions. Please refer [pgmpy#2622](pgmpy#2622) for an example of the level of detail we expect: - [x] Have you reviewed our [AI usage policy](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md#ai-usage-policy)? - [x] Are you able to fully explain your changes? We expect you to fully understand the algorithm and take full responsibility for any changes in this PR. - Please describe in **detail** how and what you used AI assistance for? Please outline your whole workflow with the AI tool. I have used claude opus 4.7 for generating parts of the PR. I first asked claude to run a line profiler to identify the most expensive operations. This pointed me towards the structure score computation and the state counts methods. This was similar to the optimization I had done for power divergence tests. I asked claude to apply a similar optimization. The tests were run to verify that the results are exactly the same as before. - What steps have you taken to verify that the changes correctly address the issue? What edge cases have you considered? Other than running tests, what else have you verified? The test results are exactly the same as before. - Have you used AI for generating tests? Can you compress them into a smaller number of tests without losing coverage? Yes, I have compressed them. ### Issue number(s) that this pull request fixes - Fixes # ### List of changes to the codebase in this pull request -
pgmpy#3383) …mpy OLS ### List of changes to the codebase in this pull request - Changes getting log-likelihood from statsmodels to doing a simple numpy OLS fit and computing it.
**The following checklist is mandatory** Your PR will be closed if you remove the checklist. Use of LLMs is **strictly forbidden** for any part of this checklist (including for improving language), and will result in a **ban** if we find any use of LLMs. ### Your checklist for this pull request - [x ] Have you followed all the steps from our [Contributing Guide](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md)? - [x ] Does the PR fully address the linked issue and is within its defined scope? If you are still working on the PR, mark it as draft. - [x ] Are all the GitHub Actions checks passing? If not, mark your PR as draft while you fix it. If you have used AI/LLMs for any assistance, please answer the following questions. Please refer [pgmpy#2622](pgmpy#2622) for an example of the level of detail we expect: - [x ] Have you reviewed our [AI usage policy](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md#ai-usage-policy)? - [x ] Are you able to fully explain your changes? We expect you to fully understand the algorithm and take full responsibility for any changes in this PR. - Please describe in **detail** how and what you used AI assistance for? Please outline your whole workflow with the AI tool. I used Antigravity to help with the heavy lifting of the migration. My workflow was pretty straightforward- I first read through the issue myself and spent some time going through the sphinxcontrib-bibtex docs to get a feel for how it works before touching anything. Once I had a clear picture, I asked Antigravity to help me write a script to do the bulk work: parsing the existing docs/references.rst, pulling out all the plaintext references, and turning them into proper BibTeX @misc entries with keyword fields for the category grouping. The same script handled scanning through the Python source files and replacing the old .. [1] Author et al. markers with :cite:p: roles. From there it was mostly me - I went through the diffs carefully, made sure each substitution actually made sense in context, and caught things like the special character issues (&, % in titles) that needed manual fixing. - What steps have you taken to verify that the changes correctly address the issue? What edge cases have you considered? Other than running tests, what else have you verified? Ran ruff check pgmpy/ - all clean. Then pytest pgmpy/tests/test_metrics/ -36/36 passed. Beyond the standard checks, I wrote a small script to make sure every :cite:p: key actually has a matching entry in references.bib. Didn't want any dangling keys slipping through across 92 files, and thankfully there were none. I also went through the BibTeX file manually for special characters since those can silently break rendering. Found 27 bare & and % in title fields and escaped them all. The test_fisher_z and test_pearsonr failures looked alarming at first but I checked git status on both files - neither was touched in this PR, so those are pre-existing and unrelated. - Have you used AI for generating tests? Can you compress them into a smaller number of tests without losing coverage? No new tests were generated. This PR is a documentation-only migration. No library logic or algorithms were changed, so no new tests are needed. ### Issue number(s) that this pull request fixes - Fixes pgmpy#3271 ### List of changes to the codebase in this pull request -Added sphinxcontrib-bibtex to the docs dependencies in pyproject.toml Turned on the sphinxcontrib.bibtex extension in docs/conf.py and pointed it to references.bib Created docs/references.bib -113 entries, each with a keyword field so they can be filtered by category Rewrote docs/references.rst to use .. bibliography:: with :filter: so the grouping is dynamic instead of hardcoded Updated 92 Python files to replace the old .. [N] Author text style with proper :cite:p: roles --------- Co-authored-by: Nitin Portfolio Builder <nitin-portfolio@example.com> Co-authored-by: Ankur Ankan <ankurankan@gmail.com>
**The following checklist is mandatory** Your PR will be closed if you remove the checklist. Use of LLMs is **strictly forbidden** for any part of this checklist (including for improving language), and will result in a **ban** if we find any use of LLMs. ### Your checklist for this pull request - [ ] Have you followed all the steps from our [Contributing Guide](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md)? - [ ] Does the PR fully address the linked issue and is within its defined scope? If you are still working on the PR, mark it as draft. - [ ] Are all the GitHub Actions checks passing? If not, mark your PR as draft while you fix it. If you have used AI/LLMs for any assistance, please answer the following questions. Please refer [pgmpy#2622](pgmpy#2622) for an example of the level of detail we expect: - [ ] Have you reviewed our [AI usage policy](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md#ai-usage-policy)? - [ ] Are you able to fully explain your changes? We expect you to fully understand the algorithm and take full responsibility for any changes in this PR. - Please describe in **detail** how and what you used AI assistance for? Please outline your whole workflow with the AI tool. [Please replace this with your answer] - What steps have you taken to verify that the changes correctly address the issue? What edge cases have you considered? Other than running tests, what else have you verified? [Please replace this with your answer] - Have you used AI for generating tests? Can you compress them into a smaller number of tests without losing coverage? [Please replace this with your answer] ### Issue number(s) that this pull request fixes - Fixes # ### List of changes to the codebase in this pull request - equal model is exist. I believe that it is occur on pgmpy#2232 PR's mistake. - https://github.com/pgmpy/pgmpy/blob/dev/pgmpy/inference/LinearModel.py - https://github.com/pgmpy/pgmpy/blob/dev/pgmpy/estimators/LinearModel.py
Added prompts for AI tool usage and workflow description.
**The following checklist is mandatory** Your PR will be closed if you remove the checklist. Use of LLMs is **strictly forbidden** for any part of this checklist (including for improving language), and will result in a **ban** if we find any use of LLMs. ### Your checklist for this pull request - [x] Have you followed all the steps from our [Contributing Guide](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md)? - [x] Does the PR fully address the linked issue and is within its defined scope? If you are still working on the PR, mark it as draft. - [ ] Are all the GitHub Actions checks passing? If not, mark your PR as draft while you fix it. If you have used AI/LLMs for any assistance, please answer the following questions. Please refer [pgmpy#2622](pgmpy#2622) for an example of the level of detail we expect: - [x] Have you reviewed our [AI usage policy](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md#ai-usage-policy)? - [x] Are you able to fully explain your changes? We expect you to fully understand the algorithm and take full responsibility for any changes in this PR. - Please describe in **detail** how and what you used AI assistance for? Please outline your whole workflow with the AI tool. - "I used an AI assistant (Antigravity) to identify the specific doctest errors (such as invalid arguments like return_type and output format mismatches). The AI also helped me run the tests locally to find the correct deterministic outputs by adding seed=42. I have manually reviewed all the changes and confirmed they are correct. "I ran the doctests locally using python -m doctest pgmpy/sampling/Sampling.py and verified that they all passed after the fixes." "No, I did not use AI for generating tests. I only fixed existing doctests." ### Issue number(s) that this pull request fixes - Fixes pgmpy#3155 ### List of changes to the codebase in this pull request -
**The following checklist is mandatory** Your PR will be closed if you remove the checklist. Use of LLMs is **strictly forbidden** for any part of this checklist (including for improving language), and will result in a **ban** if we find any use of LLMs. ### Your checklist for this pull request - [x] Have you followed all the steps from our [Contributing Guide](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md)? - [x] Does the PR fully address the linked issue and is within its defined scope? If you are still working on the PR, mark it as draft. - [x] Are all the GitHub Actions checks passing? If not, mark your PR as draft while you fix it. If you have used AI/LLMs for any assistance, please answer the following questions. Please refer [pgmpy#2622](pgmpy#2622) for an example of the level of detail we expect: - [x] Have you reviewed our [AI usage policy](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md#ai-usage-policy)? - [x] Are you able to fully explain your changes? We expect you to fully understand the algorithm and take full responsibility for any changes in this PR. - Please describe in **detail** how and what you used AI assistance for? Please outline your whole workflow with the AI tool. I have used claude code opus 4.7 for refactoring parts of the code. The original implementation was provided by @srhmm . I have refactored the code to be faster, easier to read, and improved documentation. - What steps have you taken to verify that the changes correctly address the issue? What edge cases have you considered? Other than running tests, what else have you verified? A simulation-based test case has been used to verify that the algorithm is able to recover the true graph. - Have you used AI for generating tests? Can you compress them into a smaller number of tests without losing coverage? No ### Issue number(s) that this pull request fixes - Fixes # ### List of changes to the codebase in this pull request - --------- Co-authored-by: Sarah Mameche <sarah.mameche@cispa.de> Co-authored-by: nimish <nimishpurohitn@gmail.com> Co-authored-by: Nimish Purohit <90456805+Nimish-4@users.noreply.github.com>
Updated the description of Causal Inference to focus on interventional distributions.
Auto-generated by the nightly CI workflow. Merging this will improve test split balancing. Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Ankur Ankan <ankurankan@gmail.com>
**The following checklist is mandatory** Your PR will be closed if you remove the checklist. Use of LLMs is **strictly forbidden** for any part of this checklist (including for improving language), and will result in a **ban** if we find any use of LLMs. ### Your checklist for this pull request - [ ] Have you followed all the steps from our [Contributing Guide](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md)? - [ ] Does the PR fully address the linked issue and is within its defined scope? If you are still working on the PR, mark it as draft. - [ ] Are all the GitHub Actions checks passing? If not, mark your PR as draft while you fix it. If you have used AI/LLMs for any assistance, please answer the following questions. Please refer [pgmpy#2622](pgmpy#2622) for an example of the level of detail we expect: - [ ] Have you reviewed our [AI usage policy](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md#ai-usage-policy)? - [ ] Are you able to fully explain your changes? We expect you to fully understand the algorithm and take full responsibility for any changes in this PR. - Please describe in **detail** how and what you used AI assistance for? Please outline your whole workflow with the AI tool. [Please replace this with your answer] - What steps have you taken to verify that the changes correctly address the issue? What edge cases have you considered? Other than running tests, what else have you verified? [Please replace this with your answer] - Have you used AI for generating tests? Can you compress them into a smaller number of tests without losing coverage? [Please replace this with your answer] ### Issue number(s) that this pull request fixes - Fixes # ### List of changes to the codebase in this pull request -
**The following checklist is mandatory** Your PR will be closed if you remove the checklist. Use of LLMs is **strictly forbidden** for any part of this checklist (including for improving language), and will result in a **ban** if we find any use of LLMs. ### Your checklist for this pull request - [x] Have you followed all the steps from our [Contributing Guide](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md)? - [x] Does the PR fully address the linked issue and is within its defined scope? If you are still working on the PR, mark it as draft. - [x] Are all the GitHub Actions checks passing? If not, mark your PR as draft while you fix it. If you have used AI/LLMs for any assistance, please answer the following questions. Please refer [pgmpy#2622](pgmpy#2622) for an example of the level of detail we expect: - [x] Have you reviewed our [AI usage policy](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md#ai-usage-policy)? - [x] Are you able to fully explain your changes? We expect you to fully understand the algorithm and take full responsibility for any changes in this PR. - Please describe in **detail** how and what you used AI assistance for? Please outline your whole workflow with the AI tool. I used both Claude and ChatGPT as learning and support tools while working on this task. I first went through the existing code of TreeSearch myself and tried to understand how to refractor using the examples of the `ExpertInLoop`, `ExpertKnowledge` and `GES` refractoring and what is needed to be changed to make it sklearn compatible. I refactored the TreeSearch class to make it follow a sklearn-style design instead of the old pgmpy estimator pattern. Earlier, the class stored the data inside `__init__` and used an estimate() method that directly returned the graph. I changed this by removing data from `__init__` and keeping only hyperparameters there, and moved the learning logic into a `_fit(X)` method that takes the data as input. Instead of returning the graph, the model now stores results as attributes like causal_graph_ and adjacency_matrix_. I also made sure not to modify input parameters like root_node during fitting by using a local variable, so the behavior stays consistent across multiple runs. Along with this, I updated the tests and examples to use .fit() and access results through attributes, and adjusted static method usage to be called directly from the class. I did not change implementation of the algorithm rather just copy pasted it into the new file. Also added the warning message in `pgmpy/estimators/TreeSearch.py` - What steps have you taken to verify that the changes correctly address the issue? What edge cases have you considered? Other than running tests, what else have you verified? Kept consistency in the structure by comparing it with other refractored estimators. Also copy pasted the logic from `pgmpy/estimators/TreeSearch.py` so there is minimum errors from my side which can be introduced by an AI. Also added the doctests needed and validated them. - Have you used AI for generating tests? Can you compress them into a smaller number of tests without losing coverage? Didnt use AI to generate any tests, the tests were already present in `tests/test_estimators/test_TreeSearch.py`. Earlier, the tests were creating the model with data in the constructor and calling estimate() to get the result. I changed this to first initialize the model with parameters, then call .fit(data), and access the learned graph using est.causal_graph_. ### Issue number(s) that this pull request fixes - Fixes pgmpy#2575 - Fixes pgmpy#3298 ### List of changes to the codebase in this pull request - Refractored TreeSearch Algorithm into sklearn compatible - Also refractored the tests and added them - Added the warning in `pgmpy/estimators/TreeSearch.py` - Added in doctest_modules.txt --------- Co-authored-by: Ankur Ankan <ankurankan@gmail.com>
**The following checklist is mandatory** Your PR will be closed if you remove the checklist. Use of LLMs is **strictly forbidden** for any part of this checklist (including for improving language), and will result in a **ban** if we find any use of LLMs. ### Your checklist for this pull request - [ ] Have you followed all the steps from our [Contributing Guide](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md)? - [ ] Does the PR fully address the linked issue and is within its defined scope? If you are still working on the PR, mark it as draft. - [ ] Are all the GitHub Actions checks passing? If not, mark your PR as draft while you fix it. If you have used AI/LLMs for any assistance, please answer the following questions. Please refer [pgmpy#2622](pgmpy#2622) for an example of the level of detail we expect: - [ ] Have you reviewed our [AI usage policy](https://github.com/pgmpy/pgmpy/blob/dev/Contributing.md#ai-usage-policy)? - [ ] Are you able to fully explain your changes? We expect you to fully understand the algorithm and take full responsibility for any changes in this PR. - Please list the AI tool(s) used, along with the model and its version used. [Please replace this with your answer] - Please describe in **detail** how and what you used AI assistance for? Please outline your whole workflow with the AI tool. [Please replace this with your answer] - What steps have you taken to verify that the changes correctly address the issue? What edge cases have you considered? Other than running tests, what else have you verified? [Please replace this with your answer] - Have you used AI for generating tests? Can you compress them into a smaller number of tests without losing coverage? [Please replace this with your answer] ### Issue number(s) that this pull request fixes - Fixes # ### List of changes to the codebase in this pull request -
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
6 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.