JuliaStats / NMF.jl

A Julia package for non-negative matrix factorization
Other
90 stars 34 forks source link

Fix failing test #69

Closed timholy closed 1 year ago

timholy commented 1 year ago

The greedycd algorithm gets tested for its ability to reconstruct ground-truth from offset initialization. However, it also tests the effects of regularization (sparseness priors), and with regularization the "naive ground truth" is not the actual ground truth anymore. This reduces the amplitude of the regularization coefficient to reduce the perturbation and allow the test to pass.

codecov-commenter commented 1 year ago

Codecov Report

Base: 89.70% // Head: 92.38% // Increases project coverage by +2.67% :tada:

Coverage data is based on head (0e3a793) compared to base (8629a5d). Patch has no changes to coverable lines.

Additional details and impacted files ```diff @@ Coverage Diff @@ ## master #69 +/- ## ========================================== + Coverage 89.70% 92.38% +2.67% ========================================== Files 10 11 +1 Lines 408 709 +301 ========================================== + Hits 366 655 +289 - Misses 42 54 +12 ``` | [Impacted Files](https://codecov.io/gh/JuliaStats/NMF.jl/pull/69?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=JuliaStats) | Coverage Δ | | |---|---|---| | [src/multupd.jl](https://codecov.io/gh/JuliaStats/NMF.jl/pull/69?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=JuliaStats#diff-c3JjL211bHR1cGQuamw=) | `96.59% <0.00%> (-3.41%)` | :arrow_down: | | [src/utils.jl](https://codecov.io/gh/JuliaStats/NMF.jl/pull/69?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=JuliaStats#diff-c3JjL3V0aWxzLmps) | `84.09% <0.00%> (-1.10%)` | :arrow_down: | | [src/projals.jl](https://codecov.io/gh/JuliaStats/NMF.jl/pull/69?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=JuliaStats#diff-c3JjL3Byb2phbHMuamw=) | `100.00% <0.00%> (ø)` | | | [src/greedycd.jl](https://codecov.io/gh/JuliaStats/NMF.jl/pull/69?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=JuliaStats#diff-c3JjL2dyZWVkeWNkLmps) | `100.00% <0.00%> (ø)` | | | [src/initialization.jl](https://codecov.io/gh/JuliaStats/NMF.jl/pull/69?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=JuliaStats#diff-c3JjL2luaXRpYWxpemF0aW9uLmps) | `100.00% <0.00%> (ø)` | | | [src/NMF.jl](https://codecov.io/gh/JuliaStats/NMF.jl/pull/69?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=JuliaStats#diff-c3JjL05NRi5qbA==) | `100.00% <0.00%> (ø)` | | | [src/interf.jl](https://codecov.io/gh/JuliaStats/NMF.jl/pull/69?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=JuliaStats#diff-c3JjL2ludGVyZi5qbA==) | `92.98% <0.00%> (+1.31%)` | :arrow_up: | | [src/coorddesc.jl](https://codecov.io/gh/JuliaStats/NMF.jl/pull/69?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=JuliaStats#diff-c3JjL2Nvb3JkZGVzYy5qbA==) | `100.00% <0.00%> (+3.03%)` | :arrow_up: | | [src/alspgrad.jl](https://codecov.io/gh/JuliaStats/NMF.jl/pull/69?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=JuliaStats#diff-c3JjL2Fsc3BncmFkLmps) | `85.56% <0.00%> (+7.13%)` | :arrow_up: | | [src/spa.jl](https://codecov.io/gh/JuliaStats/NMF.jl/pull/69?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=JuliaStats#diff-c3JjL3NwYS5qbA==) | `89.65% <0.00%> (+7.30%)` | :arrow_up: | | ... and [1 more](https://codecov.io/gh/JuliaStats/NMF.jl/pull/69?src=pr&el=tree-more&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=JuliaStats) | | Help us with your feedback. Take ten seconds to tell us [how you rate us](https://about.codecov.io/nps?utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=JuliaStats). Have a feature suggestion? [Share it here.](https://app.codecov.io/gh/feedback/?utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=JuliaStats)

:umbrella: View full report at Codecov.
:loudspeaker: Do you have feedback about the report comment? Let us know in this issue.