[R] akima interp results to zero with less than 10 values

[Duncan Murdoch]

The akima package has a problematic license (it doesn't allow commercial 
use), so it's been recommended that people use the interp package 
instead.  When I use interp::interp instead of akima::interp, I get 
reasonable output from your example.

So that's another reason to drop akima...

Duncan Murdoch

On 26/01/2023 9:35 a.m., PIKAL Petr wrote:
> Dear all
> 
> I have this table
>> dput(mat)
> mat <- structure(c(2, 16, 9, 2, 16, 1, 1, 4, 7, 7, 44.52, 42.8, 43.54,
> 40.26, 40.09), dim = c(5L, 3L))
> 
> And I want to calculate result for contour or image plots as I did few years
> ago.
> 
> However interp does not compute the z values and gives me zeros in z matrix.
> library(akima)
> 
>> interp(mat[,1], mat[,2], mat[, 3], nx=5, ny=5)
> $x
> [1]  2.0  5.5  9.0 12.5 16.0
> 
> $y
> [1] 1.0 2.5 4.0 5.5 7.0
> 
> $z
>       [,1] [,2] [,3] [,4] [,5]
> [1,]    0    0    0    0    0
> [2,]    0    0    0    0    0
> [3,]    0    0    0    0    0
> [4,]    0    0    0    0    0
> [5,]    0    0    0    0    0
> 
> With the example from help page if less than 10 values are used, the result
> is also zero
> interp(akima$x[1:9], akima$y[1:9], akima$z[1:9], nx=5, ny=5)
> 
> but with 10 or more values the result is correctly calculated
> interp(akima$x[1:10], akima$y[1:10], akima$z[1:10], nx=5, ny=5)
> $x
> [1]  0.0000  6.1625 12.3250 18.4875 24.6500
> 
> $y
> [1]  1.24  5.93 10.62 15.31 20.00
> 
> $z
>       [,1]     [,2]     [,3]     [,4]     [,5]
> [1,]   NA       NA       NA       NA 34.60000
> [2,]   NA       NA 27.29139 27.11807 26.60971
> [3,]   NA 19.81371 19.63614 19.12778 18.61943
> [4,]   NA 14.01443 10.66531 11.13750 10.62914
> [5,]   NA       NA       NA       NA       NA
> 
> Help page says
> x, y, and z must be the same length (execpt if x is a
> SpatialPointsDataFrame) and may contain no fewer than ***four*** points.
> 
> So my understanding was that 5 poins could be used but I am obviously wrong.
> Is it a bug in interp or in the documentation or is it my poor understanding
> of the whole matter.
> 
> Best regards
> Petr
> 
> 
> ______________________________________________
> R-help using r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.