Feature/score-type #144
Feature/score-type #144
Conversation
crates/sage/src/scoring.rs
Outdated
| @@ -156,12 +163,45 @@ fn lnfact(n: u16) -> f64 { | |||
| } | |||
| } | |||
|
|
|||
| fn log_factorial(n: u16, k: u16) -> f64 { | |||
There was a problem hiding this comment.
Why use this instead of the ln_fact function? If I wanted to calculate a precise log factorial instead of an approximation, I would already be doing so :)
There was a problem hiding this comment.
I wanted to stay true to the original implementation, but it should be equivalent to using the ln_fact function with approximation instead, replaced it accordingly.
There was a problem hiding this comment.
If there's a big difference, we can generate a pre-computed lookup table and use that.
crates/sage/src/scoring.rs
Outdated
| impl Score { | ||
| /// Calculate the hyperscore for a given PSM choosing between implementations based on `score_type` | ||
| fn hyperscore(&self, score_type: ScoreType) -> f64 { | ||
| match score_type { |
There was a problem hiding this comment.
Match expression is unnecessary here, you can simply write score_type.score(...)
|
I will work on getting this rebased and merged today. Next time, please don't merge in the master branch - makes it harder to clean up and fix later :) |
lazear
force-pushed
the
feature/score-type
branch
2 times, most recently
from
9014727
to
2e2bb14
Compare
|
Makes sense, thanks! |
This PR draft provides a slightly modified version of hyperscore as it can be found in the OpenMS repository, details in #139.