sample_lsm() does not work on NA rasters
Closed this issue · 8 comments
The sample_lsm() function does not work when the sampled area is completely covered with NAs.
library(landscapemetrics)
library(raster)
#> Loading required package: sp
# prepare data ------------------------------------------------------------
r = raster(matrix(c(rep(NA, 18), 1:18), nrow = 6))
p = raster(matrix(c(rep(1, 18), rep(2, 18)), nrow = 6))
p_sp = rasterToPolygons(p, dissolve = TRUE)
#> Loading required namespace: rgeos
plot(r)
plot(p_sp, add = TRUE, lwd = 5)
# calculate metrics -------------------------------------------------------
my_metric = sample_lsm(r, p_sp, metric = "shdi")
#> Warning in seq_len(nrow(area_patch)): first element used of 'length.out'
#> argument
#> Error in seq_len(nrow(area_patch)): argument must be coercible to non-negative integerThe error probably starts here:
landscapemetrics/R/sample_lsm.R
Lines 479 to 488 in 8a32f6d
In my opinion, the proper result should return NA in the value column, and 0 in percentage_inside.
Smaller examples are attached below.
I think the functions should return value of NA in those cases. What do you think @mhesselbarth ?
library(raster)
library(landscapemetrics)
r = raster(matrix(rep(NA, 16), nrow = 4))
r
#> class : RasterLayer
#> dimensions : 4, 4, 16 (nrow, ncol, ncell)
#> resolution : 0.25, 0.25 (x, y)
#> extent : 0, 1, 0, 1 (xmin, xmax, ymin, ymax)
#> crs : NA
#> source : memory
#> names : layer
#> values : NA, NA (min, max)
# exhibit 1 ----------------------------------------------------------------
area <- landscapemetrics:::lsm_l_ta_calc(r,
directions = 8)
#> Warning in seq_len(nrow(area_patch)): first element used of 'length.out'
#> argument
#> Error in seq_len(nrow(area_patch)): argument must be coercible to non-negative integer
# exhibit 2 ---------------------------------------------------------------
result_current_plot <- calculate_lsm(landscape = r,
verbose = verbose,
progress = FALSE)
#> Error: All raster values NA.Created on 2019-11-24 by the reprex package (v0.3.0)
Hi @mhesselbarth - it works now! The only issue is the number of warnings... Would it be possible to cut it just to the two unique warnings?
library(landscapemetrics)
library(raster)
#> Loading required package: sp
r = raster(matrix(c(rep(NA, 18), 1:18), nrow = 6))
p = raster(matrix(1:36, nrow = 6))
p_sp = rasterToPolygons(p, dissolve = TRUE)
#> Loading required namespace: rgeos
plot(r)
plot(p_sp, add = TRUE, lwd = 5)
my_metric = sample_lsm(r, p_sp, metric = "shdi")
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Please use check_landscape() to ensure the input data is valid.
#> Warning: Some of buffers extend over the landscape border. Consider decreasing
#> the size argument value.Created on 2019-11-27 by the reprex package (v0.3.0)
The problem is that the warning(s) is/are thrown by calculate_lsm() which is executed for each sample plot. One possibility would be to move the check outside calculate_lsm() and set verbose = FALSE. This would return only one warning. However, this has the problem that also ALL other warnings in calculate_lsm() are not returned e.g. if less than 3 classes are present for IJI and so on.
I don't see any workaround at the moment that doesn't include any major structural changes, unfortunately.
Related to #130
Fixed with the help of @bitbacchus idea in #130: 7eb2aa8
Interestingly, it doesn't work with reprex(), but looks nice and neat on the local console for me.
Nice, it works perfectly :-)
🥇