[vtk-developers] proposed change for vtkImplicitFunction

Will Schroeder will.schroeder at kitware.com
Thu Jan 31 12:34:20 EST 2002


Hi Michael-

Sounds good to me.

Will

At 12:30 PM 1/31/2002 +0000, Michael Halle wrote:

>Hey there,
>
>I'm writing a Perlin noise generator for vtk, and that had led me to
>wanting to make a small change to vtkImplicitFunction.
>
>Right now, subclasses implement the methods "EvaluateFunction" and
>"EvaluateGradient".  For many different functions, though, the
>gradient is a byproduct of the computation of the function value; the
>two methods needlessly require extra computation. In addition, classes
>like vtkImplicitBoolean have to do even more work because that have to
>perform an operation for each value for both functional and gradient
>evaluation.
>
>To fix this problem, I propose the addition of a virtual
>"EvaluateFunctionAndGradient" method to vtkImplicitFunction.  It will
>call EvaluateFunction and EvaluateGradient.  Subclasses can override
>the method and provide their own method that will compute both
>function and gradient simultaneously.  In addition, vtkImplicitFunction
>will get a new "FunctionValueAndGradient" method that will use this
>implementation method.
>
>No subclasses will have to be changed, but at a later point those
>that would benefit from it can be optimized to use the new combined
>method.
>
>Questions, comments?  The changes are pretty small....
>
>Thanks.
>
>                                                         --Mike
>
>
>_______________________________________________
>vtk-developers mailing list
>vtk-developers at public.kitware.com
>http://public.kitware.com/mailman/listinfo/vtk-developers




More information about the vtk-developers mailing list