[GRASS-user] 0, null values and raster algebra

GRASS GRASS user
1

Hi,

I have some troubles managing 0, null values and r.colors. I just want to extract a particular habitat category from a Global Land Cover 2000 raster into a raster of presence/absence (and NULL value for ocean and waterbodies)

This makes the following:

r.mapcalc "V7_01=if(AsiaGLC2000==7,1,0)"

So far so good, I get a raster V7_01 with null values and 1 for pixel of category 7, 0 for the others and null values stay NULL

Now, I would like to transform V7_01 into a floating point map, and then use r. neighbors for a neigborhood analysis.

r.mapcalc "V7=1.0*V7_01"

I get a raster named V7 BUT strangely for any color table, 0 value is considered NULL (transparent). The original NULL values are well kept and distinct in this raster (e.g. NULL pixels are informed as NULL and 0 pixels are 0, respectively)

Any suggestion welcome, I am stuck...

Patrick

2

Le 28/08/2012 08:17, Patrick Giraudoux a écrit :

Hi,

I have some troubles managing 0, null values and r.colors. I just want to extract a particular habitat category from a Global Land Cover 2000 raster into a raster of presence/absence (and NULL value for ocean and waterbodies)

This makes the following:

r.mapcalc "V7_01=if(AsiaGLC2000==7,1,0)"

So far so good, I get a raster V7_01 with null values and 1 for pixel of category 7, 0 for the others and null values stay NULL

Now, I would like to transform V7_01 into a floating point map, and then use r. neighbors for a neigborhood analysis.

r.mapcalc "V7=1.0*V7_01"

I get a raster named V7 BUT strangely for any color table, 0 value is considered NULL (transparent). The original NULL values are well kept and distinct in this raster (e.g. NULL pixels are informed as NULL and 0 pixels are 0, respectively)

Any suggestion welcome, I am stuck...

Patrick

Sorry to respond to myself, but it looks like I am on the wrong list. I was working in GRASS but using QGIS as interface. Using and displaying the rasters within GRASS gives results that are absolutely correct. This means that the real trouble comes from QGIS reading from GRASS. So will move the question to the relevant list.

Patrick