commons-issues mailing list archives

Site index · List index
Message view
Top
From "Gilles (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (MATH-1471) BicubicInterpolatingFunction not interpolating correctly for non discrete y value
Date Fri, 12 Oct 2018 23:36:00 GMT
```
[ https://issues.apache.org/jira/browse/MATH-1471?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16648584#comment-16648584
]

Gilles commented on MATH-1471:
------------------------------

{quote}I have also attached the unit tests to this issue.
{quote}
What you have attached is an archive of your copy of the repository, plus your compiled files
and the generated web site. :(
{quote}The pull request has been made
{quote}
Where is it (I didn't get any notification)?
Alternatively to the patch file, please copy here the link to the pull request. Thanks.

> BicubicInterpolatingFunction not interpolating correctly for non discrete y value
> ---------------------------------------------------------------------------------
>
>                 Key: MATH-1471
>                 URL: https://issues.apache.org/jira/browse/MATH-1471
>             Project: Commons Math
>          Issue Type: Bug
>    Affects Versions: 3.6.1
>         Environment: JDK 1.8.0_181
>            Reporter: Tom
>            Priority: Major
>         Attachments: ApacheCommonsMathBiInterpolationTests.zip
>
>
> Upon performing a bicubic interpolation with two point (x0, y0) and (x1, y1), the returned
bicubic interpolating function returned returns the same result for variations in the estimated
y value.
> For example, my inputs are (20, 20) and (25, 25) with f(20, 20) = 64 and f(25, 25) =
6468.
> When I get the bicubic interpolating function for this and vary the estimated x, it works
fine. For (21, 20), the function returns 730.016. When I input (20, 21), the function returns
64, which is f(20, 20). For any y value in between 20 and 25, the result is 64. This is the
case for any function for which the y estimate is different from the value on the points.
> In other instances, it is varying x values that result in the same result while varying
y estimates seem to work as expected.

--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

```
Mime
View raw message