Fix dimension ordering when converting a DataArray to DimArray #971
Conversation
Previously the code would temporarily store the dimension => lookup mapping in a Dict, but that doesn't maintain insertion order. So later on it was possible for the DimArray to be created with the dimensions in the wrong order, which would cause an exception.
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## main #971 +/- ##
==========================================
+ Coverage 83.35% 83.44% +0.09%
==========================================
Files 54 54
Lines 5154 5154
==========================================
+ Hits 4296 4301 +5
+ Misses 858 853 -5 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
|
The nightly failures seem to come from ERROR: LoadError: syntax: World age increment not at top level from other packages. |
|
It's PrecompileTools.jl, JuliaLang/PrecompileTools.jl#58 fix is on the way |
|
And Makie is breaking the docs again |
|
It's AoG this time 😭 |
should be easy to fix. I can take a look tomorrow if someone else doesn't beat me to it 😄 . |
|
Precompile issue in HTTP.jl is fixed in main revision JuliaWeb/HTTP.jl#1218 |
|
triggering CI to see if JuliaLang/PrecompileTools.jl#60 and MakieOrg/AlgebraOfGraphics.jl#646 fixed the issues. |
|
Ok it's just nightly, we can ignore that for now. Thanks everyone I'll merge an bump a patch version |
…z#971) Previously the code would temporarily store the dimension => lookup mapping in a Dict, but that doesn't maintain insertion order. So later on it was possible for the DimArray to be created with the dimensions in the wrong order, which would cause an exception.
Previously the code would temporarily store the dimension => lookup mapping in a Dict, but that doesn't maintain insertion order. So later on it was possible for the DimArray to be created with the dimensions in the wrong order, which would cause an exception.
I think it would be good to get this in a release soon-ish since it's quite a serious bug, e.g. if a square matrix is being loaded then the dimensions could get applied in the wrong order and other cases may just fail.