diff options
| author | Stefan van der Walt <stefanv@berkeley.edu> | 2018-07-25 23:32:28 -0700 |
|---|---|---|
| committer | Stefan van der Walt <stefanv@berkeley.edu> | 2018-07-25 23:32:28 -0700 |
| commit | 149636c6b764bfaf441a35115eacf0b3b1bc17f7 (patch) | |
| tree | 544e730f290487e32e1ecfb37ceea776504ce22b /doc/neps | |
| parent | a4e6f5a12db146221975a2f3d1574434d2df3bf3 (diff) | |
| download | numpy-149636c6b764bfaf441a35115eacf0b3b1bc17f7.tar.gz | |
scope: Clarify that we do care about speed
Diffstat (limited to 'doc/neps')
| -rw-r--r-- | doc/neps/scope.rst | 6 |
1 files changed, 5 insertions, 1 deletions
diff --git a/doc/neps/scope.rst b/doc/neps/scope.rst index 440f6256e..a675b8c96 100644 --- a/doc/neps/scope.rst +++ b/doc/neps/scope.rst @@ -39,4 +39,8 @@ Here, we describe aspects of N-d array computation that are within scope for Num - f2py (generating bindings for Fortran code) - testing utilities -- **Speed**: while we take performance concerns seriously, where conflict arises maintenance and portability take precedence over performance. We aim to prevent regressions where possible (e.g., through asv). +- **Speed**: we take performance concerns seriously and aim to execute + operations on large arrays with similar performance as native C + code. That said, where conflict arises, maintenance and portability take + precedence over performance. We aim to prevent regressions where + possible (e.g., through asv). |