Opened 6 years ago
Closed 6 years ago
#17633 closed defect (fixed)
Cygwin numerical noise
Reported by: | gouezel | Owned by: | |
---|---|---|---|
Priority: | major | Milestone: | sage-6.5 |
Component: | porting: Cygwin | Keywords: | |
Cc: | jpflori | Merged in: | |
Authors: | Sebastien Gouezel | Reviewers: | Jean-Pierre Flori |
Report Upstream: | N/A | Work issues: | |
Branch: | a82910b (Commits) | Commit: | a82910b360a04e59f735858d400ac85f92e8dfca |
Dependencies: | Stopgaps: |
Description (last modified by )
Many doctests fail on cygwin64 only because of numerical noise. For instance, RDF(e)
yields 2.718281828459045
on linux, but 2.7182818284590455
on cygwin.
This patch adds numerical tolerance to the failing doctests, to let them pass.
Change History (15)
comment:1 Changed 6 years ago by
- Branch set to u/gouezel/cygwin_numerical_noise
comment:2 Changed 6 years ago by
- Commit set to 91a1c8be09ab830819f9e478dfea1e6822a1b59e
comment:3 Changed 6 years ago by
- Cc jpflori added
- Component changed from PLEASE CHANGE to porting: Cygwin
- Description modified (diff)
- Status changed from new to needs_review
- Type changed from PLEASE CHANGE to defect
comment:4 Changed 6 years ago by
- Commit changed from 91a1c8be09ab830819f9e478dfea1e6822a1b59e to 63bcce1ae5d009eda8525b81f2898d9818dc8d2d
Branch pushed to git repo; I updated commit sha1. New commits:
63bcce1 | Additional noise on 6.5 beta6
|
comment:5 Changed 6 years ago by
comment:6 Changed 6 years ago by
- Commit changed from 63bcce1ae5d009eda8525b81f2898d9818dc8d2d to c4c9caa89ce153489d728bd72c987808f3805513
Branch pushed to git repo; I updated commit sha1. New commits:
c4c9caa | 17633: Replace e with log(2) in a failing doctest
|
comment:7 follow-up: ↓ 8 Changed 6 years ago by
Hi Sébastien,
From the diff two files disappeared: src/sage/functions/exp_integral.py
and src/sage/functions/hyperbolic.py
. Is that normal?
Vincent
comment:8 in reply to: ↑ 7 Changed 6 years ago by
Replying to vdelecroix:
No, it is definitely not normal. I am confused since none of the commits in the branch deletes those files. Any idea of what I might have done wrong, or is it trac diff that gets something wrong?
In any case, I can prepare a new clean branch in a few days.
comment:9 Changed 6 years ago by
- Status changed from needs_review to needs_work
comment:10 Changed 6 years ago by
- Status changed from needs_work to needs_review
No need to rebase, everything is back to normal: trac diff does not show any more that the files should be deleted.
comment:11 Changed 6 years ago by
- Status changed from needs_review to needs_work
- Work issues set to conflicts
Merge conflicts need to be fixed.
comment:12 Changed 6 years ago by
- Commit changed from c4c9caa89ce153489d728bd72c987808f3805513 to a82910b360a04e59f735858d400ac85f92e8dfca
Branch pushed to git repo; I updated commit sha1. New commits:
a82910b | #17633: remove cygwin numerical noise
|
comment:13 Changed 6 years ago by
- Status changed from needs_work to needs_review
- Work issues conflicts deleted
Rebased on 6.7 (there was no merge conflict, strangely)
comment:14 Changed 6 years ago by
- Reviewers set to Jean-Pierre Flori
- Status changed from needs_review to positive_review
Looks good.
comment:15 Changed 6 years ago by
- Branch changed from u/gouezel/cygwin_numerical_noise to a82910b360a04e59f735858d400ac85f92e8dfca
- Resolution set to fixed
- Status changed from positive_review to closed
Branch pushed to git repo; I updated commit sha1. New commits:
Some tighter bounds