Closed basaks closed 7 months ago
Merging #141 (97ae7f7) into develop (154efbc) will decrease coverage by
0.85%
. The diff coverage is33.09%
.
@@ Coverage Diff @@
## develop #141 +/- ##
===========================================
- Coverage 66.58% 65.73% -0.86%
===========================================
Files 43 43
Lines 5076 5165 +89
===========================================
+ Hits 3380 3395 +15
- Misses 1696 1770 +74
Impacted Files | Coverage Δ | |
---|---|---|
uncoverml/scripts/gridsearch_cli.py | 19.00% <6.81%> (-10.21%) |
:arrow_down: |
uncoverml/validate.py | 70.39% <14.28%> (-0.52%) |
:arrow_down: |
uncoverml/scripts/learn_cli.py | 53.02% <20.00%> (ø) |
|
uncoverml/optimise/models.py | 65.97% <44.77%> (-12.38%) |
:arrow_down: |
uncoverml/config.py | 75.81% <46.15%> (ø) |
|
uncoverml/image.py | 92.25% <100.00%> (ø) |
|
uncoverml/learn.py | 81.08% <100.00%> (-2.26%) |
:arrow_down: |
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 6729143...97ae7f7. Read the comment docs.
Based on this post: https://towardsdatascience.com/regression-prediction-intervals-with-xgboost-428e0a018b
Modifications were required due to scikit learn adaptation of the quantile functionality, and adpatation in our pipeline. Also added variance calculation assuming the upper and lower quantiles can be assumed to come from a normal distribution.