Uh oh!
There was an error while loading. Please reload this page.
Fix JSON encoding, Comparison coords, and docs warnings - #599
Merged
Conversation
FBumann
marked this pull request as ready for review
February 4, 2026 11:30
Contributor
Caution Review failedThe pull request is closed. 📝 WalkthroughWalkthroughAdds two module-private helpers to extract, merge, and re-apply non-index coordinates across xarray Datasets; refactors Comparison concatenation flow to extract non-index coords, concat with Changes
Sequence Diagram(s)(omitted) Estimated code review effort🎯 4 (Complex) | ⏱️ ~45 minutes Possibly related PRs
Poem
🚥 Pre-merge checks | ✅ 3✅ Passed checks (3 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Contributor
There was a problem hiding this comment.
Actionable comments posted: 1
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (1)
flixopt/comparison.py (1)
309-315:⚠️ Potential issue | 🟠 MajorApply coord preservation pattern to
solutionandinputsproperties.Both
solution(line 309) andinputs(line 378) usexr.concat()withcoords='minimal', which drops non-index coords (e.g.,componentoncontributordim). This causes loss of label coordinates that are needed in the combined result.The helper functions
_extract_nonindex_coords()and_apply_merged_coords()already exist and are used successfully in_concat_property()for stats concatenation. Apply the same pattern here:Suggested pattern
- self._solution = xr.concat(- [ds.expand_dims(case=[name]) for ds, name in zip(datasets, self._names, strict=True)],- dim='case',- join='outer',- coords='minimal',- fill_value=float('nan'),- )+ expanded = [ds.expand_dims(case=[name]) for ds, name in zip(datasets, self._names, strict=True)]+ expanded, merged_coords = _extract_nonindex_coords(expanded)+ result = xr.concat(expanded, dim='case', join='outer', coords='minimal', fill_value=float('nan'))+ self._solution = _apply_merged_coords(result, merged_coords)- self._inputs = xr.concat(- [ds.expand_dims(case=[name]) for ds, name in zip(datasets, self._names, strict=True)],- dim='case',- join='outer',- coords='minimal',- fill_value=float('nan'),- )+ expanded = [ds.expand_dims(case=[name]) for ds, name in zip(datasets, self._names, strict=True)]+ expanded, merged_coords = _extract_nonindex_coords(expanded)+ result = xr.concat(expanded, dim='case', join='outer', coords='minimal', fill_value=float('nan'))+ self._inputs = _apply_merged_coords(result, merged_coords)
🤖 Fix all issues with AI agents
In `@flixopt/comparison.py`:
- Around line 58-59: When merging coord mappings in comparison.py, detect
conflicts where the same dim value (dv) already exists in merged[name][1] but
maps to a different coord value (cv); currently the code silently keeps the
first value (merged[name][1][dv] = cv). Modify the logic around the
merged[name][1] assignment to check if dv in merged[name][1] and
merged[name][1][dv] != cv, and then either raise a ValueError (including
identifiers like name, dv, existing value and new cv) or emit a clear warning
(using the project logger) so callers are alerted to inconsistent inputs instead
of silently keeping the first mapping.
Uh oh!
There was an error while loading. Please reload this page.
- Apply _extract_nonindex_coords pattern to solution and inputs properties - Add warning when coordinate mappings conflict during merge Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
…comparison # Conflicts: # CHANGELOG.md
Contributor
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Fix all issues with AI agents
In `@flixopt/comparison.py`:
- Around line 74-85: The _apply_merged_coords function can produce None entries
because mapping.get(dv) returns None for values added by outer joins; change the
logic that builds new_coords for each (name, (dim, mapping)) so unmapped
dimension values keep a sensible fallback (e.g., use the original dv or a
sentinel) instead of None — in practice replace mapping.get(dv) with
mapping.get(dv, dv) (or an explicit sentinel) when creating the list for
new_coords[name], ensuring ds.coords[dim].values are mapped safely before
calling ds.assign_coords(new_coords).
🧹 Nitpick comments (1)
flixopt/comparison.py (1)
32-71: Implementation looks solid with conflict detection in place.The logic for identifying non-index coordinates (single-dim coords where dim ≠ name) and merging their mappings is correct. The conflict warning at lines 60-65 properly addresses the previous review feedback.
One minor consideration:
stacklevel=4assumes a specific call depth. Since this helper is called from multiple sites (solution,inputs,_concat_property,_combine_data), the stack depth varies. The warning will point to different locations depending on the caller, which may confuse users. Consider usingstacklevel=2to point to the helper itself, or documenting which call site triggered the warning in the message.💡 Optional: Include caller context in warning message
elif merged[name][1][dv] != cv: warnings.warn( f"Coordinate '{name}' has conflicting values for dim value '{dv}': " - f"'{merged[name][1][dv]}' vs '{cv}'. Keeping first value.",- stacklevel=4,+ f"'{merged[name][1][dv]}' vs '{cv}'. Keeping first value.",+ stacklevel=2, )
Uh oh!
There was an error while loading. Please reload this page.
…ne for unmapped values. Change it to mapping.get(dv, dv) so unmapped dimension values fall back to themselves. ⏺ Update(flixopt/comparison.py) ⎿ Added 1 line, removed 1 line 80 for name, (dim, mapping) in merged.items(): 81 if dim not in ds.dims: 82 continue 83 - new_coords[name] = (dim, [mapping.get(dv) for dv in ds.coords[dim].values]) 83 + new_coords[name] = (dim, [mapping.get(dv, dv) for dv in ds.coords[dim].values]) 84 85 return ds.assign_coords(new_coords) 86 ⏺ Done. The change on line 83 ensures that when mapping doesn't contain a key for a dimension value (which happens with outer-join additions), the original value dv is preserved instead of inserting None.
Contributor
There was a problem hiding this comment.
Actionable comments posted: 1
🤖 Fix all issues with AI agents
In `@flixopt/comparison.py`:
- Around line 32-71: The function _extract_nonindex_coords currently builds a
global coords_to_drop set which can remove a coord that is an index in some
datasets; change it to track non-index coords per dataset (e.g., build a
list/sequence of per-dataset sets or a dict keyed by dataset index) while still
accumulating merged mappings only for coords that are non-index in that
particular dataset; then when dropping, iterate datasets and call
ds.drop_vars(per_dataset_coords_to_drop[i], errors='ignore') for each dataset so
indexed coords in other datasets are preserved; keep using merged (dict[str,
tuple[str, dict]]) and the same logic for filling merged[name][1] and warnings.
Comment on lines
+32
to
+71
| def _extract_nonindex_coords(datasets: list[xr.Dataset]) -> tuple[list[xr.Dataset], dict[str, tuple[str, dict]]]: | ||
| """Extract and merge non-index coords, returning cleaned datasets and merged mappings. | ||
| Non-index coords (like `component` on `contributor` dim) cause concat conflicts. | ||
| This extracts them, merges the mappings, and returns datasets without them. | ||
| """ | ||
| if not datasets: | ||
| return datasets, {} | ||
| # Find non-index coords and collect mappings | ||
| merged: dict[str, tuple[str, dict]] = {} | ||
| coords_to_drop: set[str] = set() | ||
| for ds in datasets: | ||
| for name, coord in ds.coords.items(): | ||
| if len(coord.dims) != 1: | ||
| continue | ||
| dim = coord.dims[0] | ||
| if dim == name or dim not in ds.coords: | ||
| continue | ||
| coords_to_drop.add(name) | ||
| if name not in merged: | ||
| merged[name] = (dim, {}) | ||
| for dv, cv in zip(ds.coords[dim].values, coord.values, strict=False): | ||
| if dv not in merged[name][1]: | ||
| merged[name][1][dv] = cv | ||
| elif merged[name][1][dv] != cv: | ||
| warnings.warn( | ||
| f"Coordinate '{name}' has conflicting values for dim value '{dv}': " | ||
| f"'{merged[name][1][dv]}' vs '{cv}'. Keeping first value.", | ||
| stacklevel=4, | ||
| ) | ||
| # Drop these coords from datasets | ||
| if coords_to_drop: | ||
| datasets = [ds.drop_vars(coords_to_drop, errors='ignore') for ds in datasets] | ||
| return datasets, merged |
Contributor
There was a problem hiding this comment.
🧩 Analysis chain
🏁 Script executed:
# First, verify the file content and examine the functioncd /
git ls-files | grep -E "flixopt/comparison\.py"Repository: flixOpt/flixopt
Length of output: 127
🏁 Script executed:
# Read the actual code in the file
cat -n flixopt/comparison.py | sed -n '32,71p'Repository: flixOpt/flixopt
Length of output: 1925
🏁 Script executed:
# Look for tests or usage examples of this function
fd -e py | xargs rg "_extract_nonindex_coords" -B 2 -A 5Repository: flixOpt/flixopt
Length of output: 3092
🏁 Script executed:
# Search for test files related to comparison or this function
fd -e py test| xargs rg "extract_nonindex_coords|_extract_nonindex_coords"2>/dev/null ||echo"No direct test found"Repository: flixOpt/flixopt
Length of output: 79
🏁 Script executed:
# Search for any tests in the repo that might test this scenario
fd -e py -path "*/test*"| head -20Repository: flixOpt/flixopt
Length of output: 227
🏁 Script executed:
# Look for any comments or TODOs about index coords
rg "index.*coord|coord.*index" flixopt/comparison.py -i -B 2 -A 2Repository: flixOpt/flixopt
Length of output: 2025
🏁 Script executed:
# Check if there are any GitHub issues or comments mentioning this
rg "coords_to_drop" --include="*.py" -B 3 -A 3Repository: flixOpt/flixopt
Length of output: 500
🏁 Script executed:
# Search for test files
find . -name "*test*.py" -type f 2>/dev/null | head -20Repository: flixOpt/flixopt
Length of output: 800
🏁 Script executed:
# Search for xarray usage and index behavior
rg "drop_vars|\.index" flixopt/comparison.py -B 1 -A 1Repository: flixOpt/flixopt
Length of output: 446
🏁 Script executed:
# Check xarray documentation or behavior - look for how index coords work in xarray
rg "is_index|IndexVariable". -l 2>/dev/null | head -5Repository: flixOpt/flixopt
Length of output: 41
🏁 Script executed:
# Look for any coords that are explicitly indexed
rg "\.set_index|\.reset_index". -B 2 -A 2 2>/dev/null | head -20Repository: flixOpt/flixopt
Length of output: 1242
🏁 Script executed:
# Check if there are any tests specifically for comparison or concat operations
find tests -name "*.py" -type f -exec grep -l "comparison\|_extract_nonindex_coords\|_concat" {} \;Repository: flixOpt/flixopt
Length of output: 303
🏁 Script executed:
# Look at how datasets are constructed and used in the function calls
rg "expand_dims.*case" flixopt/comparison.py -B 5 -A 5Repository: flixOpt/flixopt
Length of output: 2331
🏁 Script executed:
# Check xarray's behavior - look for any documentation or comments about index coords
rg "dim == name" flixopt/ -B 3 -A 3Repository: flixOpt/flixopt
Length of output: 421
🏁 Script executed:
# Examine the test_comparison.py file to understand test coverage
head -100 tests/test_comparison.pyRepository: flixOpt/flixopt
Length of output: 2966
🏁 Script executed:
# Check for any tests that use _extract_nonindex_coords or test the indexing behavior
rg "_extract_nonindex_coords\|index.*coord" tests/test_comparison.py -B 3 -A 3Repository: flixOpt/flixopt
Length of output: 41
🏁 Script executed:
# Look at the full test_comparison.py file to understand what scenarios are tested
wc -l tests/test_comparison.pyRepository: flixOpt/flixopt
Length of output: 87
🏁 Script executed:
# Search for tests that might exercise the index coord scenario
rg "coords|index|concat" tests/test_comparison.py -i -B 2 -A 2 | head -50Repository: flixOpt/flixopt
Length of output: 1542
🏁 Script executed:
# Let me look for the actual test functions to see if any test concat or coords
rg "^def test_" tests/test_comparison.pyRepository: flixOpt/flixopt
Length of output: 41
🏁 Script executed:
# Check if xarray's drop_vars handles index coords specially
python3 << 'EOF'import systry: import xarray as xr import numpy as np # Create a test scenario: dataset with an index coordinate ds1 = xr.Dataset({ 'var': (['x', 'y'], np.random.rand(3, 4)) }, coords={ 'x': [0, 1, 2], 'y': [10, 20, 30, 40], 'component': (['y'], ['a', 'b', 'c', 'd']) # non-index coord on y }) # Set y as an index ds1 = ds1.set_index(y='y') print("Dataset 1:") print(ds1) print("\nCoords in ds1:") for name, coord in ds1.coords.items(): print(f" {name}: dims={coord.dims}, is_index={name in ds1.indexes}") # Now create ds2 where 'component' is an index coord on a different dim ds2 = xr.Dataset({ 'var': (['component', 'z'], np.random.rand(4, 5)) }, coords={ 'component': ['a', 'b', 'c', 'd'], 'z': [100, 200, 300, 400, 500] }) ds2 = ds2.set_index(component='component') print("\n\nDataset 2:") print(ds2) print("\nCoords in ds2:") for name, coord in ds2.coords.items(): print(f" {name}: dims={coord.dims}, is_index={name in ds2.indexes}") # Test: Try to drop 'component' from ds2 (which is indexed) print("\n\nAttempting to drop 'component' from ds2 with errors='ignore'...") ds2_dropped = ds2.drop_vars('component', errors='ignore') print("Result:") print(ds2_dropped) print("\nCoords after drop:") for name, coord in ds2_dropped.coords.items(): print(f" {name}: dims={coord.dims}, is_index={name in ds2_dropped.indexes}")except Exception as e: print(f"Error: {e}", file=sys.stderr) import traceback traceback.print_exc()EOFRepository: flixOpt/flixopt
Length of output: 209
🏁 Script executed:
# Look at how the datasets are being used after _extract_nonindex_coords
rg "_apply_merged_coords" flixopt/comparison.py -B 2 -A 5Repository: flixOpt/flixopt
Length of output: 1945
🏁 Script executed:
# Search for the _apply_merged_coords function to understand what it does
rg "def _apply_merged_coords" flixopt/comparison.py -A 15Repository: flixOpt/flixopt
Length of output: 647
🏁 Script executed:
# Check if there are any comments or documentation about why the condition is `dim == name`
rg "dim == name|index.*coord" flixopt/ -B 5 -A 5Repository: flixOpt/flixopt
Length of output: 10071
🏁 Script executed:
# Search online documentation or look for any issues/PRs related to this
rg "index.*coord.*drop|drop.*index". -i --max-count 10Repository: flixOpt/flixopt
Length of output: 178
🏁 Script executed:
# Search for any xarray-related issues or discussions in comments
rg "xarray|drop_vars|index" flixopt/comparison.py -B 1 -A 1 | head -30Repository: flixOpt/flixopt
Length of output: 1182
🏁 Script executed:
# Check if there's any handling or workaround for index coords elsewhere in the codebase
rg "drop_vars" flixopt/ -B 2 -A 2Repository: flixOpt/flixopt
Length of output: 5735
Track non-index coords per-dataset instead of globally to preserve index coordinates.
coords_to_drop accumulates globally across datasets, but a coord name can be indexed in one dataset and non-indexed in another. This causes indexed coords to be dropped unintentionally. For example, if "x" is a non-index coord on dim "y" in dataset A and an index coord on dim "x" in dataset B, it gets added to the global set from A, then dropped from all datasets including B where it's actually indexed.
Drop only those coords that are non-index in each dataset:
🔧 Proposed fix
# Find non-index coords and collect mappings
merged: dict[str, tuple[str, dict]] = {}
coords_to_drop: set[str] = set()
for ds in datasets:
+ ds_drop: set[str] = set()
for name, coord in ds.coords.items():
if len(coord.dims) != 1:
continue
dim = coord.dims[0]
if dim == name or dim not in ds.coords:
continue
coords_to_drop.add(name)
+ ds_drop.add(name)
if name not in merged:
merged[name] = (dim, {})
for dv, cv in zip(ds.coords[dim].values, coord.values, strict=False):
if dv not in merged[name][1]:
merged[name][1][dv] = cv
elif merged[name][1][dv] != cv:
warnings.warn(
f"Coordinate '{name}' has conflicting values for dim value '{dv}': "
f"'{merged[name][1][dv]}' vs '{cv}'. Keeping first value.",
stacklevel=4,
)
+ coords_to_drop_list.append(ds_drop) if 'coords_to_drop_list' in dir() else None+ if 'coords_to_drop_list' not in dir():+ coords_to_drop_list = [ds_drop]
# Drop these coords from datasets
if coords_to_drop:
- datasets = [ds.drop_vars(coords_to_drop, errors='ignore') for ds in datasets]+ datasets = [+ ds.drop_vars(drop, errors='ignore')+ for ds, drop in zip(datasets, coords_to_drop_list, strict=True)+ ]🤖 Prompt for AI Agents
In `@flixopt/comparison.py` around lines 32 - 71, The function
_extract_nonindex_coords currently builds a global coords_to_drop set which can
remove a coord that is an index in some datasets; change it to track non-index
coords per dataset (e.g., build a list/sequence of per-dataset sets or a dict
keyed by dataset index) while still accumulating merged mappings only for coords
that are non-index in that particular dataset; then when dropping, iterate
datasets and call ds.drop_vars(per_dataset_coords_to_drop[i], errors='ignore')
for each dataset so indexed coords in other datasets are preserved; keep using
merged (dict[str, tuple[str, dict]]) and the same logic for filling
merged[name][1] and warnings.
Uh oh!
There was an error while loading. Please reload this page.
FBumann added a commit
that referenced
this pull request
Feb 5, 2026
* Update the CHANGELOG.md * Update to tsam v3.1.0 and add warnings for preserve_n_clusters=False * [ci] prepare release v6.0.0 * fix typo in deps * fix typo in README.md * Revert citation temporarily * [ci] prepare release v6.0.0 * Improve json io * fix: Notebooks using tsam * Allow manual docs dispatch * Created: tests/test_clustering/test_multiperiod_extremes.py Test Coverage (56 tests): Multi-Period with Different Time Series - TestMultiPeriodDifferentTimeSeries - Tests for systems where each period has distinct demand profiles: - Different cluster assignments per period - Optimization with period-specific profiles - Correct expansion mapping per period - Statistics correctness per period Extreme Cluster Configurations - TestExtremeConfigNewCluster - Tests method='new_cluster': - Captures peak demand days - Can increase cluster count - Works with min_value parameter - TestExtremeConfigReplace - Tests method='replace': - Maintains requested cluster count - Works with multi-period systems - TestExtremeConfigAppend - Tests method='append': - Combined with segmentation - Objective preserved after expansion Combined Multi-Period and Extremes - TestExtremeConfigMultiPeriod - Extremes with multi-period/scenario: - Requires preserve_n_clusters=True for multi-period - Works with periods and scenarios together - TestMultiPeriodWithExtremes - Combined scenarios: - Different profiles with extreme capture - Extremes combined with segmentation - Independent cluster assignments per period Multi-Scenario Clustering - TestMultiScenarioWithClustering - Scenarios with clustering - TestFullDimensionalClustering - Full (periods + scenarios) combinations IO Round-Trip - TestMultiPeriodClusteringIO - Save/load preservation tests Edge Cases - TestEdgeCases - Single cluster, many clusters, occurrence sums, mapping validation * fix: clustering and tsam 3.1.0 issue * [ci] prepare release v6.0.1 * fix: clustering and tsam 3.1.0 issue * [ci] prepare release v6.0.1 * ci: remove test * [ci] prepare release v6.0.1 * chore(deps): update dependency werkzeug to v3.1.5 (#564) * chore(deps): update dependency ruff to v0.14.14 (#563) * chore(deps): update dependency netcdf4 to >=1.6.1, <1.7.5 (#583) * chore(deps): update dependency pre-commit to v4.5.1 (#532) * fix: Comparison coords (#599) * Fix coords concat in comparison.py * Fix coords concat in comparison.py * Fix coords concat in comparison.py * Add 6.0.1 changelog entry Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com> * Fix coord preservation in Comparison.solution and .inputs - Apply _extract_nonindex_coords pattern to solution and inputs properties - Add warning when coordinate mappings conflict during merge Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com> * Update CHANGELOG.md * Update CHANGELOG.md * ⏺ The fix is straightforward — on line 83, mapping.get(dv) returns None for unmapped values. Change it to mapping.get(dv, dv) so unmapped dimension values fall back to themselves. ⏺ Update(flixopt/comparison.py) ⎿ Added 1 line, removed 1 line 80 for name, (dim, mapping) in merged.items(): 81 if dim not in ds.dims: 82 continue 83 - new_coords[name] = (dim, [mapping.get(dv) for dv in ds.coords[dim].values]) 83 + new_coords[name] = (dim, [mapping.get(dv, dv) for dv in ds.coords[dim].values]) 84 85 return ds.assign_coords(new_coords) 86 ⏺ Done. The change on line 83 ensures that when mapping doesn't contain a key for a dimension value (which happens with outer-join additions), the original value dv is preserved instead of inserting None. * Update Changelog --------- Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com> * [ci] prepare release v6.0.2 * typo * Revert "typo" This reverts commit 4a57282. * Add plan file * Add comprehensive test_math coverage for multi-period, scenarios, clustering, and validation - Add 26 new tests across 8 files (×3 optimize modes = ~75 test runs) - Multi-period: period weights, flow_hours limits, effect limits, linked invest, custom period weights - Scenarios: scenario weights, independent sizes, independent flow rates - Clustering: basic objective, storage cyclic/intercluster modes, status cyclic mode - Storage: relative min/max charge state, relative min/max final charge state, balanced invest - Components: transmission startup cost, Power2Heat, HeatPumpWithSource, SourceAndSink - Flow status: max_uptime standalone test - Validation: SourceAndSink requires size with prevent_simultaneous * ⏺ Done. Here's a summary of what was changed: Fix (flixopt/components.py:1146-1169): In _relative_charge_state_bounds, the scalar else branches now expand the base parameter to regular timesteps only (timesteps_extra[:-1]), then concat with the final-timestep DataArray containing the correct override value. Previously they just broadcast the scalar across all timesteps, silently ignoring relative_minimum_final_charge_state / relative_maximum_final_charge_state. Tests (tests/test_math/test_storage.py): Added two new tests — test_storage_relative_minimum_final_charge_state_scalar and test_storage_relative_maximum_final_charge_state_scalar — identical scenarios to the existing array-based tests but using scalar defaults (the previously buggy path). * Added TestClusteringExact class with 3 tests asserting exact per-timestep values in clustered systems: 1. test_flow_rates_match_demand_per_cluster — Verifies Grid flow_rate matches demand [10,20,30,40] identically in each cluster, objective = 200. 2. test_per_timestep_effects_with_varying_price — Verifies per-timestep costs [10,20,30,40] reflect price×flow with varying prices [1,2,3,4] and constant demand=10, objective = 200. 3. test_storage_cyclic_charge_discharge_pattern — Verifies storage with cyclic clustering: charges at cheap timesteps (price=1), discharges at expensive ones (price=100), with exact charge_state trajectory across both clusters, objective = 100. Deviation from plan: Used equal cluster weights [1.0, 1.0] instead of [1.0, 2.0]/[1.0, 3.0] for tests 1 and 2. This was necessary because cluster_weight is not preserved during NetCDF roundtrip (pre-existing IO bug), which would cause the save->reload->solve mode to fail. Equal weights produce correct results in all 3 IO modes while still testing the essential per-timestep value correctness. * More storage tests * Add multi-period tests * Add clustering tests and fix issues with user set cluster weights * Update CHANGELOG.md * Mark old tests as stale * Update CHANGELOG.md * Mark tests as stale and move to new dir * Move more tests to stale * Change fixtures to speed up tests * Moved files into stale * Renamed folder * Reorganize test dir * Reorganize test dir * Rename marker * 2. 08d-clustering-multiperiod.ipynb (cell 29): Removed stray <cell_type>markdown</cell_type> from Summary cell 3. 08f-clustering-segmentation.ipynb (cell 33): Removed stray <cell_type>markdown</cell_type> from API Reference cell 4. flixopt/comparison.py: _extract_nonindex_coords now detects when the same coord name appears on different dims — warns and skips instead of silently overwriting 5. test_multiperiod_extremes.py: Added .item() to mapping.min()/.max() and period_mapping.min()/.max() to extract scalars before comparison 6. test_flow_status.py: Tightened test_max_uptime_standalone assertion from > 50.0 to assert_allclose(..., 60.0, rtol=1e-5) matching the documented arithmetic --------- Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> Co-authored-by: renovate[bot] <29139614+renovate[bot]@users.noreply.github.com> Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
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 freeto join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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.
Summary
Patch release v6.0.1 with bug fixes:
ensure_ascii=Falsetojson.dumps()calls inio.pycomponentcoordinate becoming(case, contributor)shaped after concatenation inComparisonclass. Non-index coordinates are now properly merged before concatpreserve_n_clusters=Trueto allExtremeConfigcalls to fix FutureWarning from tsam v3.1workflow_dispatchinputs for manual docs deployment with version selectionTest plan
Comparison.flow_hourshas correctcomponentcoordinate shape🤖 Generated with Claude Code
Summary by CodeRabbit
Bug Fixes
Documentation