Repository navigation
Read the numeric fields of records in bulk - #123
Merged
Merged
Conversation
ramdump read every field of a record through an RNTupleView, one call per value; in a profile of a full read, RFieldBase::Read alone took 7.5% of the CPU time. The fifteen numeric fields are now read with bulk reads into vectors, a run of rows of a cluster at a time, starting at 1,024 rows and doubling to 65,536 while the rows come in order, so a small region reads as many pages as before. On chr20 of NA12878 a full read takes 16.8 s instead of 18.1 s.
|
clang-tidy review says "All clean, LGTM! 👍" |
Codecov Report❌ Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## develop #123 +/- ##
===========================================
+ Coverage 89.00% 89.02% +0.02%
===========================================
Files 26 26
Lines 3364 3408 +44
Branches 1127 1155 +28
===========================================
+ Hits 2994 3034 +40
Misses 196 196
- Partials 174 178 +4
Flags with carried forward coverage won't be shown. Click here to find out more.
🚀 New features to boost your workflow:
|
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
ramdump'sRecordReaderread every field of a record through anRNTupleView, one callper value. In a perf profile of a full read of chr20 (29.7 s of CPU),
RFieldBase::Readtook 7.5% and most of the 8.3% of the per-row callback.
The fifteen numeric fields (FLAG, reference, position, MAPQ, mate reference, the mate
deltas, NM, AS, XS, MQ, RG, PG, the tag order and the compression flags) are now read with
RNTupleView::CreateBulk()into vectors, a run of rows of one cluster at a time; for thesesimple fields a bulk read copies whole pages. A run starts at 1,024 rows and doubles up to
65,536 while the rows come in order, so a small region reads as many pages as before. The
strings and vectors are read as before.
NA12878 chr20 at 30x (17.7 million records) on a Xeon E5-2690 v4, pinned to one socket,
files in the page cache, medians of 5 runs in alternating order, against #119:
-h)The output is the same, byte for byte, for every step and for regions and flag filters on
2 million records of chromosome 11. All tests pass. No format change.