Closed darsnack closed 2 years ago
Merging #59 (c0a19e3) into main (e1f4c6b) will increase coverage by
0.21%
. The diff coverage is75.00%
.
@@ Coverage Diff @@
## main #59 +/- ##
==========================================
+ Coverage 88.91% 89.13% +0.21%
==========================================
Files 13 13
Lines 415 414 -1
==========================================
Hits 369 369
+ Misses 46 45 -1
Impacted Files | Coverage Δ | |
---|---|---|
src/obstransform.jl | 84.09% <57.14%> (ø) |
|
src/batchview.jl | 76.92% <100.00%> (+0.60%) |
:arrow_up: |
src/observation.jl | 86.95% <100.00%> (+1.53%) |
:arrow_up: |
src/obsview.jl | 65.78% <100.00%> (ø) |
Continue to review full report at Codecov.
Legend - Click here to learn more
Δ = absolute <relative> (impact)
,ø = not affected
,? = missing data
Powered by Codecov. Last update e1f4c6b...c0a19e3. Read the comment docs.
I support this change 👍 Makes things simpler for newcomers, makes it easier to take advantage of types from other packages that already implement indexing, and one no longer needs to import (and export) getobs
and numobs
everywhere. Also the same interface as torch.utils.data.Dataset
and familiarity is good.
This removes the
getindex
/length
defaults fromAbstractDataContainer
. Instead, types implementing the interface should implementBase.getindex
orBase.length
first, and only implementgetobs
/numobs
when the fallback definitions don't apply.Fixes #57.