Re: Bug in Interpolation for multi-d data?
- To: mathgroup at smc.vnet.net
- Subject: [mg86188] Re: Bug in Interpolation for multi-d data?
- From: Antti Penttilä@smc.vnet.net
- Date: Wed, 5 Mar 2008 03:40:17 -0500 (EST)
- Organization: University of Helsinki
- References: <fq8r74$jf4$1@smc.vnet.net> <fqb863$moo$1@smc.vnet.net> <fqetcl$jj6$1@smc.vnet.net> <fqghfg$f53$1@smc.vnet.net>
Antti Penttil=E4 wrote:
>
> It is a bug in at least 6.0.0 and 6.0.1. I hope that it is fixed in
> 6.0.2, since Wolfram knows about this.
I confirm that it seems to be corrected in V6.0.2.
You can test the bug with e.g. following code where 3D interpolation is
made for ordered and non-onredered lists and 2D plots are drawn. In
6.0.1, the latter plot produces wrong result.
t = Flatten[
Table[ {i, j, i (j + RandomReal[{-2, 2}])}, {i, 10}, {j, 10}], 1];
int1 = Interpolation[ t]
Show[
Plot[ int1[ x, 5], {x, 1, 10}],
ListPlot[ Cases[t, x_ /; x[[2]] == 5][[All, {1, 3}]]]
]
t1 = t[[RandomSample[ Range[Length[ t]]]]];
int2 = Interpolation[ t1]
Show[
Plot[ int2[ x, 5], {x, 1, 10}],
ListPlot[ Cases[t1, x_ /; x[[2]] == 5][[All, {1, 3}]]]
]
Antti