Closed reedacartwright closed 7 years ago
Merging #225 into develop will decrease coverage by
0.42%
. The diff coverage is83%
.
@@ Coverage Diff @@
## develop #225 +/- ##
===========================================
- Coverage 67.99% 67.57% -0.43%
===========================================
Files 68 68
Lines 6384 6440 +56
Branches 1123 1126 +3
===========================================
+ Hits 4341 4352 +11
- Misses 1576 1630 +54
+ Partials 467 458 -9
Impacted Files | Coverage Δ | |
---|---|---|
src/include/dng/probability.h | 100% <ø> (ø) |
:arrow_up: |
src/include/dng/depths.h | 94.07% <ø> (ø) |
:arrow_up: |
src/lib/ad.cc | 67.66% <ø> (ø) |
:arrow_up: |
src/lib/task/loglike.cc | 0% <0%> (ø) |
:arrow_up: |
src/lib/probability.cc | 46.57% <100%> (ø) |
:arrow_up: |
src/include/dng/task/call.xmh | 100% <100%> (ø) |
:arrow_up: |
src/include/dng/utility.h | 88% <100%> (+4.32%) |
:arrow_up: |
src/include/dng/genotyper.h | 100% <100%> (+13.33%) |
:arrow_up: |
src/include/dng/task/loglike.xmh | 100% <100%> (ø) |
:arrow_up: |
src/lib/task/call.cc | 81.3% <100%> (+0.32%) |
:arrow_up: |
... and 5 more |
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 83115a1...cb52670. Read the comment docs.
Using a mixture of 2 DM for genotype likelihoods ended up being too complicated from the standpoint of the end user. The
gamma
parameter was too complicated. The solution is to change the model to a single dirichlet multinomial, which allows the command line parameters to be simplified.The benefit is that calculating genotype likelihoods becomes faster.
While creating unit tests, bugs in the calculation of haploid genotype likelihoods were discovered and fixed. Other tests were adjusted to reflect the new genotype likelihood model and default parameters.
This pull request also includes improvements to some additional unit tests that began on another branch, but needed to be merged into this one.