Fix build failure by relaxing overly restrictive dependency constraints#207
Open
Copilot wants to merge 5 commits into
Open
Fix build failure by relaxing overly restrictive dependency constraints#207Copilot wants to merge 5 commits into
Copilot wants to merge 5 commits into
Conversation
- Expand Python support from 3.9-3.10 to 3.9-3.11 - Remove upper version caps on JAX and jaxlib - Update llvmlite constraint to allow newer versions - Fix numpy constraint to avoid numpy 2.0 - These changes allow the project to build on Python 3.9, 3.10, and 3.11 Co-authored-by: gordonkoehn <10867778+gordonkoehn@users.noreply.github.com>
Copilot
AI
changed the title
[WIP] Investigate and resolve build failure issue
Fix build failure by relaxing overly restrictive dependency constraints
Nov 11, 2025
Collaborator
|
@copilot now we are getting erros with JAX: please take a look at this. |
JAX 0.4.24+ removed jax.random.PRNGKeyArray which breaks the existing code. Restoring the <0.4.24 upper bound to maintain API compatibility. Co-authored-by: gordonkoehn <10867778+gordonkoehn@users.noreply.github.com>
Author
... Fixed in commit 0f8d19f. The issue was caused by removing the upper bound on JAX. JAX 0.4.24+ introduced a breaking API change that removed |
Collaborator
|
@copilot now the type checking fails: |
- Add type: ignore comments for pandas DataFrame initialization (pandas stubs are overly strict about list types for columns/index) - Fix method signature mismatches by using _IntegerTreeRoot type alias instead of anytree.Node - Fix JSONEncoder.default parameter name to match base class (obj -> o) - Add type: ignore for networkx dag_longest_path_length (accepts Graph in practice but typed as DiGraph) These errors surfaced due to updated dependency versions with stricter type stubs. Co-authored-by: gordonkoehn <10867778+gordonkoehn@users.noreply.github.com>
Author
... Fixed in commit 548f384. The type checking errors were caused by stricter type stubs in the updated dependencies (particularly pandas and its type annotations). I've fixed all 10 errors:
These were pre-existing type issues that surfaced due to the relaxed numpy constraint allowing newer pandas type stubs to be installed. |
Collaborator
|
@copilot build still fails right ? |
The # type: ignore[arg-type] comments added in the previous commit made two lines exceed black's 88-char limit, causing the black --check step to fail: - _scphylo_utils.py line 69 was 113 chars - _huntress.py line 57 was 89 chars Fix by splitting DataFrame constructor arguments onto separate lines so each line stays under the 88-char limit. Agent-Logs-Url: https://github.com/cbg-ethz/PYggdrasil/sessions/d4aa1c37-6c8b-47cf-a624-c741778957df Co-authored-by: gordonkoehn <10867778+gordonkoehn@users.noreply.github.com>
Author
Yes, it was still failing. The root cause was that the
Fixed in |
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
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.
Root Cause:
The build was failing due to overly restrictive dependency version constraints in
pyproject.toml:>=3.9,<3.11(only 3.9 and 3.10)<0.4.24) that prevented newer compatible versions^0.39.1) which was too restrictive<1.29.0) that should be<2.0.0Solution Implemented:
Updated
pyproject.tomlwith relaxed but safe version constraints:>=3.9,<3.12(supports 3.9, 3.10, and 3.11)>=0.4.4,<0.4.24(restored upper bound due to API breaking changes in 0.4.24+)>=0.4.4,<0.4.24(restored upper bound for consistency with JAX)>=0.39.1(changed from caret to >= for flexibility)<2.0.0(proper upper bound to avoid numpy 2.0 breaking changes)Type Checking Fixes:
Fixed type checking errors that surfaced due to stricter type stubs in updated dependencies:
type: ignorecomments for pandas DataFrame initialization (pandas stubs overly strict about list types)_IntegerTreeRootinstead ofanytree.Node)type: ignorefor networkx function that accepts Graph in practiceJAX Compatibility Note:
JAX 0.4.24+ introduced breaking API changes that removed
jax.random.PRNGKeyArray. The upper bound<0.4.24is necessary to maintain compatibility with the existing codebase without code changes.Limitations:
scphylo-toolsGit dependency using deprecated Python C APIOriginal prompt
💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.