Use float32 when lossless for float64 #365
Closed
Conversation
This file contains 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
When a float64 can be fully represented as a float32, use that instead. Encoding is backwards compatible with existing reader. `AppendFloat64Only` is made available to unconditionally write a float64. Micro-benchmarks are not great: ``` BenchmarkAppendFloat64Only-32 511061935 2.324 ns/op 3442.55 MB/s 0 B/op 0 allocs/op BenchmarkAppendFloat64-32 136697270 8.704 ns/op 919.08 MB/s 0 B/op 0 allocs/op ``` But microbenchmarks are micro-benchmarks. We take a much bigger cost on `AppendInt64` and other integer functions. We could make it opt-in, but that would make it pretty much unused.
klauspost
added a commit
to klauspost/msgp
that referenced
this pull request
Sep 29, 2024
Add `//msgp:compactfloats` file directive, that will store float64 as float32, if it can be done so losslessly. Boring, but correct replacement of tinylib#365
|
Replaced by #366 |
klauspost
added a commit
that referenced
this pull request
Sep 30, 2024
Add `//msgp:compactfloats` file directive, that will store float64 as float32, if it can be done so losslessly. Boring, but correct replacement of #365
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Outdated suggestions cannot be applied.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.
When a float64 can be fully represented as a float32, use that instead.
Encoding is backwards compatible with existing reader.
AppendFloat64Onlyis made available to unconditionally write a float64.Micro-benchmarks are not great:
But microbenchmarks are micro-benchmarks. We take a much bigger cost on
AppendInt64and other integer functions.We could make it opt-in, but that would make it pretty much unused.