codecov-commenter
commented
3 years ago Codecov Report
Merging #114 (019ca8b) into main (0d42a1a) will decrease coverage by
2.56%. The diff coverage is15.55%.
@@ Coverage Diff @@
## main #114 +/- ##
==========================================
- Coverage 96.90% 94.34% -2.57%
==========================================
Files 29 29
Lines 1391 1432 +41
==========================================
+ Hits 1348 1351 +3
- Misses 43 81 +38 | Impacted Files | Coverage Δ | |
|---|---|---|
| bycycle/cyclepoints/extrema.py | 58.76% <15.55%> (-37.67%) |
:arrow_down: |
Continue to review full report at Codecov.
Legend - Click here to learn more
Δ = absolute <relative> (impact),ø = not affected,? = missing dataPowered by Codecov. Last update 0d42a1a...019ca8b. Read the comment docs.
This adds an option to have redefine troughs on a burst-by-burst basis (assuming a fixed cycle duration), providing a solution to periods fluctuating +/- 10% of the simulated length when an oscillation is combined with noise. Outside of simulations, I don't think this update will be useful since a fixed cycle duration can't be assumed.
What led me here was poor fits when transforming motifs to individual cycles (i.e. when a simulated cycle has 50 samples, but bycycle defines the cycle as 45 samples, and transforming a 50 sample motif to the cycle leads to a poor fit, typically at the edge that is over/under defined). This is shown in the image below - this example has a tiny amount of noise that still throws off the cyclepoint definitions a little.