[ 
https://issues.apache.org/jira/browse/MATH-1650?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17941258#comment-17941258
 ] 

Gilles Sadowski commented on MATH-1650:
---------------------------------------

New class added in commit 82070d092e617ab5e5d58ef5f93b3df0df50d61b (in 
"master").

However, as noted previously, there seems to be quite some common code in 
{{SplineInterpolator}} and {{ClampedSplineInterpolator}}.  It would be useful, 
both for maintenance and for usage, if this could be factored out.
Also, the design of the {{interpolation}} package starts to break down with the 
method introduced in {{ClampedSplineInterpolator}} that is not part of the 
{{UnivariateInterpolator}} interface. :-(

Hence keeping this report open for now.

> Add clamped spline interpolation
> --------------------------------
>
>                 Key: MATH-1650
>                 URL: https://issues.apache.org/jira/browse/MATH-1650
>             Project: Commons Math
>          Issue Type: New Feature
>    Affects Versions: 4.X
>            Reporter: Michael Scholz
>            Priority: Minor
>              Labels: Polynomials, interpolation, spline
>         Attachments: 2022-10-05_ClampedSplineInterpolator.patch, 
> 2025-01-02_ClampedSplineInterpolator.patch
>
>
> We would like to contribute a new _clamped_ spline interpolation function in 
> addition to the already available unclamped spline function. Our new 
> {{ClampedSplineInterpolator}} is based on the same textbook as the original 
> {{{}SplineInterpolator{}}}. The clamped spline offers additional 
> parameterisation of starting and ending slopes (1st derivatives) as boundary 
> conditions in order to provide more flexibility in spline creation.
> In this patch we follow the approach of subclassing the original 
> {{SplineInterpolator}} and simply overloading it's {{interpolate()}} function 
> by these two additional parameters. Is this an acceptable way or does the 
> community recommend a different design approach?
> After clarifying the basic implementation approach we could also supply 
> necessary tests etc. and finally contribute everything via ordinary GitHub 
> pull request.
> Refer to our post on the dev mailing list: 
> https://lists.apache.org/thread/34qnx4tgjbyv345lgmd57g0bnlnwdzc8
>  



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

Reply via email to