Every commit scored for change-risk against this repo's own history, so 'elevated' means elevated here rather than on some global curve.
Needs review
302 commits sit in this repo's top risk tercile, which is 33% of the 916scored. The cut is drawn against this codebase's own history rather than a global curve, so a quiet repo still fills its top band, and here it starts at 4.3 out of 10. What pushes a commit up is size and spread together: a large change confined to one area scores below a smaller one scattered across a dozen files.
Commit categories over time, read off the subject line. Fixes carry the accent because that is the series this chart exists to show.
Began other-led, now leaning test.
Ranked by change-risk, highest first. Priority is a tercile of this repo's own distribution, so a quiet repo still fills its top band.
| # | Commit | Author | When | Lines | Review priority | Top driver |
|---|---|---|---|---|---|---|
| 1 | f49f0058Added initial doc test | Jake Lever | 9y ago | +14.9K -0 | 100thElevated | more lines added than baseline |
| 2 | 2fa3b139Change doc theme | Jake Lever | 9y ago | +1.8K -270 | 100thElevated | more lines added than baseline |
| 3 | 410d0babAdded in vectorizer code from VERSE (to be refactored and integrated) | Jake Lever | 9y ago | +1.1K -0 | 100thElevated | more lines added than baseline |
| 4 | 33a9e704Added BioC loading code into kindred (for now) | Jake Lever | 9y ago | +710 -0 | 100thElevated | more lines added than baseline |
| 5 | a2ebe083Update tests for new Spacy model, with focus on vectorizer | Jake Lever | 7y ago | +772 -197 | 100thElevated | more lines added than baseline |
| 6 | 4aef0ec2Starting to build EntityRecognizer code inside Kindred | Jake Lever | 8y ago | +451 -0 | 99thElevated | more lines added than baseline |
| 7 | cdadc3d1Saving some old debug code | Jake Lever | 9y ago | +378 -0 | 99thElevated | more lines added than baseline |
| 8 | e2c85416Parser tests now work | Jake Lever | 9y ago | +428 -4 | 99thElevated | more lines added than baseline |
| 9 | 88a5ec87Cleanup of init and separated datatypes in separate files | Jake Lever | 9y ago | +324 -330 | 99thElevated | more lines added than baseline |
| 10 | 33677134Restructured tests into more files | Jake Lever | 9y ago | +309 -278 | 99thElevated | more lines added than baseline |
| 11 | 693ee804Added function for saving corpus to a csv file. | vinitjain | 5y ago | +235 -183 | 99thElevated | more lines added than baseline |
| 12 | e958899dMega-update: Relations can now be n-ary, not just binary | Jake Lever | 8y ago | +351 -62 | 99thElevated | more lines added than baseline |
| 13 | 19073d0dAdded non-English language code (#2) | Jake Lever | 9y ago | +332 -8 | 99thElevated | more lines added than baseline |
| 14 | c7925e0fStart of major refactor: RelationData -> Corpus/Document | Jake Lever | 9y ago | +212 -214 | 98thElevated | more lines added than baseline |
| 15 | 6e323671Added some results data | Jake Lever | 9y ago | +204 -0 | 98thElevated | more lines added than baseline |
| 16 | 6aa8e1c2Adding ST loading | Jake Lever | 9y ago | +232 -11 | 98thElevated | more lines added than baseline |
| 17 | 39bb585cRefactor vectorizer for new candidate relation list input | Jake Lever | 8y ago | +218 -151 | 98thElevated | more lines added than baseline |
| 18 | d837a944Cleaner code for loading simple tag | Jake Lever | 9y ago | +230 -10 | 98thElevated | more lines added than baseline |
| 19 | 2942a0b9Added VERSE vectorizer and it works | Jake Lever | 9y ago | +205 -9 | 98thElevated | more lines added than baseline |
| 20 | b64ac0ceAdd documentation on file formats (#3) | Jake Lever | 7y ago | +173 -8 | 98thElevated | more lines added than baseline |
| 21 | 515484b9Update docs for Corpus & Document | Jake Lever | 9y ago | +166 -61 | 98thElevated | more lines added than baseline |
| 22 | f76eb242More code for paper (not really tests at the moments) | Jake Lever | 9y ago | +149 -0 | 98thElevated | more lines added than baseline |
| 23 | 773964d0Basic relation candidate tests passed | Jake Lever | 9y ago | +177 -114 | 98thElevated | more lines added than baseline |
| 24 | 047173f4Change multiclass behaviour to multilabels using separate classifier for each rel type, allows overlap | Jake Lever | 7y ago | +180 -58 | 97thElevated | more lines added than baseline |
| 25 | 2e0e8efeAdded EntityRecognizer to package properly with initial test | Jake Lever | 8y ago | +229 -293 | 97thElevated | more lines added than baseline |
| 26 | ab7c3e56Updates of docs for several classes | Jake Lever | 9y ago | +158 -32 | 97thElevated | more lines added than baseline |
| 27 | 76d61be7BioC loading test works | Jake Lever | 9y ago | +168 -5 | 97thElevated | more lines added than baseline |
| 28 | 90dad12cMake convertTaggedText test work | Jake Lever | 9y ago | +161 -7 | 97thElevated | more lines added than baseline |
| 29 | fd65ca0cImprove how temp directories are used in tests | Jake Lever | 7y ago | +132 -127 | 97thElevated | more lines added than baseline |
| 30 | 5d851d36Refactor tests for Spacy integration | Jake Lever | 8y ago | +134 -246 | 97thElevated | more lines added than baseline |
| 31 | 51539df9Added sklearn SVM to kindred | Jake Lever | 9y ago | +143 -53 | 97thElevated | more lines added than baseline |
| 32 | b1707411Attempt at api doc | Jake Lever | 9y ago | +112 -0 | 97thElevated | more lines added than baseline |
| 33 | 0d1e6234Tests for merge, acronyms and wordlist load in EntityRecognizer | Jake Lever | 8y ago | +167 -3 | 96thElevated | more lines added than baseline |
| 34 | 77984382EntityRecognizer tests for fusions and term merges | Jake Lever | 8y ago | +139 -2 | 96thElevated | more lines added than baseline |
| 35 | 67a3f9e6Add experimental code using Spacy | Jake Lever | 8y ago | +145 -0 | 96thElevated | more lines added than baseline |
| 36 | 9f6f1936Added extra testing for TFIDF in vectorizer | Jake Lever | 9y ago | +141 -1 | 96thElevated | more lines added than baseline |
| 37 | 1311fa9dAdded rename of bionlpst tools | Jake Lever | 9y ago | +131 -0 | 96thElevated | more lines added than baseline |
| 38 | 2a02705aFix merging of multiple entities | Jake Lever | 5y ago | +140 -53 | 96thElevated | more lines added than baseline |
| 39 | d739deaeAdded variant and polymorphism extraction with tests | Jake Lever | 8y ago | +136 -1 | 96thElevated | more lines added than baseline |
| 40 | e73db669Big change to embed candidate relations in Sentence objects and relationtypes in Corpus | Jake Lever | 9y ago | +120 -79 | 96thElevated | more lines added than baseline |
| 41 | 61d645d2Gave classifier some options for single/multi classifiers and feature building | Jake Lever | 9y ago | +117 -52 | 96thElevated | more lines added than baseline |
| 42 | 86e11b25Add PubAnnotation format to output and sourceRelationID to relation | Jake Lever | 6y ago | +122 -26 | 95thElevated | more lines added than baseline |
| 43 | ad55c975Simplify load functions down to a single one | Jake Lever | 7y ago | +165 -174 | 95thElevated | more lines added than baseline |
| 44 | aed9302eUpdates to docs | Jake Lever | 9y ago | +110 -20 | 95thElevated | more lines added than baseline |
| 45 | 8ce36000Add first draft of tutorial code | Jake Lever | 7y ago | +102 -0 | 95thElevated | more lines added than baseline |
| 46 | 3580bb2aAdd microRNA tests | Jake Lever | 8y ago | +109 -1 | 95thElevated | more lines added than baseline |
| 47 | 5c9e179bFix Vectorizer so it can be pickled | Jake Lever | 9y ago | +103 -101 | 95thElevated | more lines added than baseline |
| 48 | c68af9a6Many more tests for RelationClassifier | Jake Lever | 9y ago | +141 -5 | 95thElevated | more lines added than baseline |
| 49 | 09b67f8fIntegrated PubAnnotation | Jake Lever | 9y ago | +116 -41 | 95thElevated | more lines added than baseline |
| 50 | 8711104cAdded BioNLP tests for all 2016 BB3 and SeeDev relation datasets | Jake Lever | 9y ago | +127 -11 | 95thElevated | more lines added than baseline |
Two views of the same model: where the cuts fall, and what commit shape lands you above them.
Every scored commit, binned on the raw 0 to 10 score rather than the percentile. Percentile ranks are uniform by construction, so that axis has no shape to draw. The dashed lines are the tercile cuts behind each row's priority pill.
The 200 most recent commits, on their own recency sample rather than the feed above: that defaults to risk-sorted, so reusing it would plot only the top tercile and call it the spread. Big and scattered is what the model penalises. Click a dot to open it.
KGerring/kindred has 928 commits in its history from 2 contributors, the first of them Apr 14, 2017. No file in the repository changed in the last 90 days. Every commit is scored for change risk from its size, spread and the history of the files it touches.
