[Date Index] [Thread Index] [Author Index]

ListDensityPlot with interpolation: strictly non-negative data

• To: mathgroup at smc.vnet.net
• Subject: [mg91807] ListDensityPlot with interpolation: strictly non-negative data
• From: Szabolcs <szhorvat at gmail.com>
• Date: Tue, 9 Sep 2008 06:53:53 -0400 (EDT)

Hello,

I have some 2D scalar data on a grid that I would like to plot with
higher-than-first order interpolation.  ListDensityPlot can do this
with the setting InterpolationOrder -> 3.

The problem in this specific case is that the data is strictly non-
negative, but the interpolation creates negative regions.  Note that
clipping at 0 is not a good solution because it creates a non-smooth
surface and, even worse, the interpolated function might have positive
"ripples" in it. (I.e. it can decay to 0 by first becoming negative,
then positive again, then 0.)

Is there a simple way to plot strictly non-negative data with higher
order interpolation (perhaps an existing package that implements a
special interpolation)?