[GRASS-dev] r.colors -a rule sorting

GRASS GRASS developer
1

Hi,

I am just trying out the new 'r.colors -a'. Results look good, visually
matches 'r.mapcalc abs($MAP)' + 'r.colors -g' exactly.

One thing I notice in the colr/ file, the positive/negative rules are interleaved:

====
nv:255
*:255
1:0:0:255 1.1457610996:0:7:255
-1.1457610996:0:7:255 -1:0:0:255
1.1457610996:0:7:255 1.3127684973:0:15:255
-1.3127684973:0:15:255 -1.1457610996:0:7:255
1.3127684973:0:15:255 1.5041190769:0:22:255
...
-618328.3661061861:255:16:0 -539666.049353256:255:23:0
618328.3661061861:255:16:0 708456.5886435417:255:8:0
-708456.5886435417:255:8:0 -618328.3661061861:255:16:0
708456.5886435417:255:8:0 811721:255:0:0
-811721:255:0:0 -708456.5886435417:255:8:0

(this is r.watershed accumulation map for spearfish's elevation.10m)

is it any problem that they are not sorted min to max, or is it just
cosmetic?

Hamish

2

Hamish wrote:

I am just trying out the new 'r.colors -a'. Results look good, visually
matches 'r.mapcalc abs($MAP)' + 'r.colors -g' exactly.

In general, it won't be an exact match, as -a uses "log(abs(x)+1)" so
that values close to zero don't "hog" the colour table. If your range
is small, you will probably notice the difference (if your range
includes values very close to zero, you would definitely notice).

One thing I notice in the colr/ file, the positive/negative rules are interleaved:

====
nv:255
*:255
1:0:0:255 1.1457610996:0:7:255
-1.1457610996:0:7:255 -1:0:0:255
1.1457610996:0:7:255 1.3127684973:0:15:255
-1.3127684973:0:15:255 -1.1457610996:0:7:255
1.3127684973:0:15:255 1.5041190769:0:22:255
...
-618328.3661061861:255:16:0 -539666.049353256:255:23:0
618328.3661061861:255:16:0 708456.5886435417:255:8:0
-708456.5886435417:255:8:0 -618328.3661061861:255:16:0
708456.5886435417:255:8:0 811721:255:0:0
-811721:255:0:0 -708456.5886435417:255:8:0

(this is r.watershed accumulation map for spearfish's elevation.10m)

is it any problem that they are not sorted min to max, or is it just
cosmetic?

Just cosmetic.

--
Glynn Clements <glynn@gclements.plus.com>