Skip to content

Commit 2fe69e6

Browse files
committedOct 18, 2014
Merge pull request #1635 from gioman/saga_slope
Fixes SAGA slope,aspect,curvature for SAGA 2.1.x
2 parents d77e0f0 + 28f3f46 commit 2fe69e6

File tree

2 files changed

+18
-0
lines changed

2 files changed

+18
-0
lines changed
 
Lines changed: 18 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,18 @@
1+
Slope, Aspect, Curvature
2+
ta_morphometry
3+
ParameterRaster|ELEVATION|Elevation|False
4+
ParameterSelection|METHOD|Method|[0] Maximum Slope (Travis et al. 1975);[1] Maximum Triangle Slope (Tarboton 1997);[2] Least Squares Fitted Plane (Horn 1981, Costa-Cabral & Burgess 1996);[3] 6 parameter 2nd order polynom (Evans 1979);[4] 6 parameter 2nd order polynom (Heerdegen & Beran 1982);[5] 6 parameter 2nd order polynom (Bauer, Rohdenburg, Bork 1985);[6] 9 parameter 2nd order polynom (Zevenbergen & Thorne 1987);[7]10 parameter 3rd order polynom (Haralick 1983)|6
5+
ParameterSelection|UNIT_SLOPE|Slope Units|[0] radians;[1] degree;[2] percent|0
6+
ParameterSelection|UNIT_ASPECT|Aspect Units|[0] radians;[1] degree|0
7+
OutputRaster|SLOPE|Slope
8+
OutputRaster|ASPECT|Aspect
9+
OutputRaster|C_GENE|General Curvature
10+
OutputRaster|C_PLAN|Plan Curvature
11+
OutputRaster|C_PROF|Profile Curvature
12+
OutputRaster|C_TANG|Tangential Curvature
13+
OutputRaster|C_LONG|Longitudinal Curvature
14+
OutputRaster|C_CROS|Cross-Sectional Curvature
15+
OutputRaster|C_MINI|Minimal Curvature
16+
OutputRaster|C_MAXI|Maximal Curvature
17+
OutputRaster|C_TOTA|Total Curvature
18+
OutputRaster|C_ROTO|Flow-Line Curvature

5 commit comments

Comments
 (5)

volaya commented on Oct 19, 2014

@volaya
Contributor

The issue with this algorithm was related to saga 2.1.2.

This commit will make the algorithm work fine with 2.1.2, but will break it if the user has 2.1.0 installed, which was working fine before.

We still do not have a mechanism to handle multiple version of SAGA (only the distinction between 2.0.8 and another one), so this kind of fixes should wait until that is implemented.

This commit causes a regression if using 2.1.0, so it should be reverted

gioman commented on Oct 19, 2014

@gioman
Contributor

Hi,
it seems to me that modules descriptions renamed 2.1_* are working also with SAGA 2.1.2 so maybe the idea could be the following: Processing supports SAGA 2.0.8 and that's fine as it isshipped with osgeo4w, and then supports always the latest SAGA release, as it is usually available for download for Windows and in Linux repositories like Ubuntugis (used by most to install QGIS anyway).

volaya commented on Oct 19, 2014

@volaya
Contributor

hmm, sounds like a good idea, to have a reference version which is older (2.0.8 in this case), and then support the latest one to avoid being blocked in case you distribution only has that.

It 's ok, but we should document it. For instance, with this latest change, 2.1.0 will not work for the slope algorithm, so you need whether 2.0.8 or the latest version, not an intermediate on

gioman commented on Oct 19, 2014

@gioman
Contributor

So I would suggest something along this lines:

Processing in QGIS 2.6 supports two SAGA versions: 2.0.8 and the latest currently available, 2.1.2

On Windows QGIS ships SAGA 2.0.8, and Processing is configured to use it. If you want to use SAGA 2.1.2 then you must install QGIS with the Osgeo4W installer (and just it) and then download/install SAGA from this link http://sourceforge.net/projects/saga-gis/files/SAGA%20-%202.1/SAGA%202.1.2/

After that you must configure QGIS/Processing to use SAGA 2.1.2 by entering Processing general options.

On Ubuntu/Debian Linux you will find SAGA 2.1.2 in Ubuntugis and Debiangis repositories, the same usually used to install QGIS.

pcav commented on Oct 19, 2014

@pcav
MemberAuthor

In the longer term, IMHO it would be better if the backend version would be an attribute of each module, so one could have as many modules as he wants, and processing would pick up the suitable one.

Please sign in to comment.