Conversation
Codecov Report❌ Patch coverage is Additional details and impacted files@@ Coverage Diff @@
## master #400 +/- ##
=======================================
Coverage 73.00% 73.00%
=======================================
Files 439 439
Lines 36672 36685 +13
Branches 5036 5036
=======================================
+ Hits 26772 26782 +10
- Misses 8785 8788 +3
Partials 1115 1115 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
ddaspit
left a comment
There was a problem hiding this comment.
@ddaspit reviewed 5 files and all commit messages, and made 1 comment.
Reviewable status:complete! all files reviewed, all discussions resolved (waiting on Enkidu93).
Enkidu93
left a comment
There was a problem hiding this comment.
@ddaspit, just wanted to make sure you saw my question above. I'm assuming you're OK with the -1.0 default then?
@Enkidu93 made 1 comment.
Reviewable status:complete! all files reviewed, all discussions resolved (waiting on Enkidu93).
ddaspit
left a comment
There was a problem hiding this comment.
I think using the geometric mean of the word confidences would be good.
@ddaspit made 1 comment.
Reviewable status:complete! all files reviewed, all discussions resolved (waiting on Enkidu93).
Enkidu93
left a comment
There was a problem hiding this comment.
Done 👍. Since we're now using the geometric mean in a few places, I've moved it to a utility class. Let me know if it ought to be somewhere else or if you'd prefer I not do that.
@Enkidu93 made 1 comment.
Reviewable status: 3 of 7 files reviewed, all discussions resolved (waiting on ddaspit).
ddaspit
left a comment
There was a problem hiding this comment.
@ddaspit reviewed 4 files and all commit messages, and made 1 comment.
Reviewable status: all files reviewed, 1 unresolved discussion (waiting on Enkidu93).
src/SIL.Machine/Utils/ConfidenceHelper.cs line 14 at r2 (raw file):
/// <param name="values"></param> /// <returns>The geometric mean.</returns> public static double GeometricMean(IList<double> values)
Please move this to StatisticalMethods.
Enkidu93
left a comment
There was a problem hiding this comment.
@Enkidu93 made 1 comment.
Reviewable status: 3 of 8 files reviewed, 1 unresolved discussion (waiting on ddaspit).
src/SIL.Machine/Utils/ConfidenceHelper.cs line 14 at r2 (raw file):
Previously, ddaspit (Damien Daspit) wrote…
Please move this to
StatisticalMethods.
Done. A perfect spot that I didn't know existed haha. Interestingly none of the other methods in this class seem to be used anywhere in the codebase 🤔. Just curious: I guess these are more for external use then?
ddaspit
left a comment
There was a problem hiding this comment.
@ddaspit reviewed 5 files and all commit messages, made 2 comments, and resolved 1 discussion.
Reviewable status:complete! all files reviewed, all discussions resolved (waiting on Enkidu93).
src/SIL.Machine/Utils/ConfidenceHelper.cs line 14 at r2 (raw file):
Previously, Enkidu93 (Eli C. Lowry) wrote…
Done. A perfect spot that I didn't know existed haha. Interestingly none of the other methods in this class seem to be used anywhere in the codebase 🤔. Just curious: I guess these are more for external use then?
Yes, these are available for external use.
Many changes were not relevant to Machine (C#). My only question: For the
TranslationEngines that create translation results in Machine, should I just provide a dummy -1.0 value for the sequence confidence or should I combine the individual confidences in some way to create the sequence confidence (e.g. geometric mean)?Fixes #396.
Port changes from sillsdev/machine.py#279.
This change is