Fix rationalize(x) accuracy with tiny floats / when x is a rational
#60828
Conversation
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
- Fix `tol` ignored when `T` is not specified. - Specific implementation for rationals (vs floats): return `x` unless the given `tol` is significant enough to reduce the magnitude of its components, in which case rationalize `float(x)` instead (see JuliaLang#60768).
Convert `x` to a float type of higher precision when needed, so that `a, r = divrem(x,y)` remain exact in the while loop (this adds type instability and breaks inference so the function return type needs to be specified).
Comment on lines
+256
to
+260
| if p > precision(Float64) | ||
| return setprecision(() -> rationalize(T, BigFloat(x), tol), BigFloat, p) | ||
| end | ||
| x, a, r = Float64.((x, a, r)) | ||
| end |
Member
There was a problem hiding this comment.
This will break type stability.
Author
There was a problem hiding this comment.
The conversion depends on the value of x, so yes. I made the function return type explicit to preserve inference.
I don't think there is a way to fix the inaccuracy issues without giving x more precision.
Author
There was a problem hiding this comment.
Maybe we should use a wrapper function that does the check and conversion if needed, so that the current (inner) function remains type stable ?
Co-authored-by: Oscar Smith <oscardssmith@gmail.com>
Co-authored-by: Oscar Smith <oscardssmith@gmail.com>
nsajko
added
rationals
Co-authored-by: Oscar Smith <oscardssmith@gmail.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 participants
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Outdated suggestions cannot be applied.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.
Fixes #60768, fixes #49803, fixes #49848.
I made it in a single PR because there isn't much code involved and because one fix needs the other. More details in the commit messages and issue descriptions.