1

I am having troubles interpretating the output of r.cost.full.

Currently i am working on a project comparing some arcgis tools with qgis tools (workflow and output). One of my examples is to make a cost distance raster with both softwares.

The input Cost Raster and the Input Start Points (point features) are the same in both softwares. My problem here is that the result values of arcgis are within 0 and ~7 which make sense in terms of interpretating. But the result values of r.cost.full are between ~0.02 and ~0.3. When i visualize the output rasters with the same amount of classes, the rasters look very similar.

I ran the calculations on arcmap 10.2 and 10.3 and on qgis 2.6 and 2.8.2

Does anyone know why the values are so different? Am i missing something obvious here?

1

I know this question has been asked for a long time now, but maybe somebody will find my findings interesting. I solved it also a long time ago, but simply forgot to answer here. Sorry for that.

Anyway the difference between ArcMap and QGIS (r.cost.full) lies within the algorithm. While ArcMap multiplies the costs automatically with the cell size, r.cost.full doesn't.

In order to get comparable results, i had to multiply the output raster in QGIS with its cell size. The comparison then showed minor differences. I figured that ArcMap and QGIS use different lengths of real numbers, meaning the difference come from rounding methods.

0

The QGIS Processing toolbox is very well done, but the problem is that the "external" applications used (GRASS GIS, SAGA GIS, ...) are not QGIS and most QGIS users are unfamiliar with them.

The original GRASS GIS command is r.cost (6.x) or r.cost (7.x). This is a command with many options.

enter image description here

They are simplified in the toolbox with 3 commands, r.cost, r.cost.full and r.cost.raster with less options.

I think that before comparing the results with ArcGIS, you need to know first the usage of the command in GRASS GIS, or understand and verify if the GRASS command parameters are the same as the ArcGIS command parameters, to apply it with profit in QGIS

  • I am aware of the toolbox mechanism adding functions from GRASS GIS and SAGA GIS. And as you said, it is indeed very well done. As far as i found out, the command parameters (if they are the input parameters) are used the same way in both algorithms. As an output both algorithms create raster, where each cell represents the least cost from one or more given starting points. – rabear Jun 13 '15 at 7:28
  • Ok, but the signification of the results ? – gene Jun 13 '15 at 8:13
  • The result should be a raster, where the pixel value represents the time needed to pass it. – rabear Jun 13 '15 at 8:21
  • I don't know ArcGIS but in r.cost diagonal costs are multiplied by a factor that depends on the dimensions of the cell – gene Jun 13 '15 at 9:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.