[
https://issues.apache.org/jira/browse/NUMBERS-86?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16727325#comment-16727325
]
Gilles commented on NUMBERS-86:
-------------------------------
{quote}It seems cleaner to me to have the call to positivePolarForm in a single
location
{quote}
In most cases, I'd have agreed with you; but here the "end" is treated
specially, and this fact is blurred by a "universal fix".
Reading the code, one also wonders why the {{apply}} from the {{Linear}} and
{{Spherical}} algos don't ensure positive polar form. Of course "technically",
it is obvious that it's done later, but conceptually, is it correct that they
can return a quaternion that is not in positive polar form? [(Very) contrived
case is that those classes could be reused later.]
> Slerp Unit Tests
> ----------------
>
> Key: NUMBERS-86
> URL: https://issues.apache.org/jira/browse/NUMBERS-86
> Project: Commons Numbers
> Issue Type: Test
> Reporter: Matt Juntunen
> Priority: Major
> Time Spent: 20m
> Remaining Estimate: 0h
>
> {{Slerp}} needs unit tests. There are tests in commons-geometry but they rely
> on {{Vector3D}}. We need to port those tests or write new ones.
>
> Pull request: https://github.com/apache/commons-numbers/pull/23
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
