Access is denied - r.statistics

[nkale]

Related to ticket #1038 - Running the following:

r.statistics --verbose base=HUC@PERMANENT cover=ag@PERMANENT method=sum output=huc_ag

results in the following error:

Access is denied. ERROR: o_sum: running r.stats command

Same results when attempting to run an average as well. The fix mentioned in the other ticket (change security settings to full control for USER) didn't work.

[hamish]

does r.stats run on its own? how about g.tempfile pid=1234?

note o_sum.c in gr65 calls r.stats via system() with:

#define STATS "r.stats"
#define RECLASS "r.reclass"
...
    sprintf(command, "%s -cn input=\"%s,%s\" fs=space > %s", STATS, basemap,
            covermap, tempfile1);

    if (stat = system(command)) {
        unlink(tempfile1);
        G_fatal_error(_("%s: running %s command"), me, STATS);
    }

but before dealing with system(), a starting point might be:

Index: o_sum.c
===================================================================
--- o_sum.c     (revision 42566)
+++ o_sum.c     (working copy)
@@ -27,7 +27,7 @@
     tempfile1 = G_tempfile();
     tempfile2 = G_tempfile();
 
-    sprintf(command, "%s -cn input=\"%s,%s\" fs=space > \"%s\"", STATS, basemap,
+    sprintf(command, "%s -cn input=\"%s,%s\" fs=space output=\"%s\"", STATS, basemap,
            covermap, tempfile1);
 
     if (stat = system(command)) {

Hamish

[nkale]

Didn't really follow anything below "note 0_sum.c in gr65...", but as for the commands you suggested:

r.stats ag@PERMANENT - runs fine, no errors, meaningful output r.stats HUC@PERMANENT - runs fine, no errors, meaningful output g.tempfile pid=1234 - no errors, output is:

C:/Users/Nat Kale/Documents/Grass_GIS DataBase/Alberts_Landcover/PERMANENT/.tmp/1234.1

That is the correct location & mapset. Can't think of anything else to try.

[hamish]

could you give the grass 7 daily snapshot a try? it uses an improved method to generate the stats and hopefully bypasses the problem.

system() replaced with G_system() in grass 6.5svn with r42570, daily WinGrass snapshots for 6.5 are probably broken until bug #1089 gets fixed, and it this wants testing there before backporting to 6.4.

cheers, Hamish

[nkale]

Replying to hamish:

could you give the grass 7 daily snapshot a try? it uses an improved method to generate the stats and hopefully bypasses the problem.

system() replaced with G_system() in grass 6.5svn with r42570, daily WinGrass snapshots for 6.5 are probably broken until bug #1089 gets fixed, and it this wants testing there before backporting to 6.4.

cheers, Hamish

That worked. Ran with a slightly smaller dataset (didn't know if the raster format was the same between 6 & 7, so I had to do some lengthy re-importing), but I didn't get any errors, and the output map made sense.

On the other hand, in 7 I can no longer launch GRASS in MSYS mode, but that's maybe a bug for another ticket.

Thanks,

Nat

[hamish]

Replying to hamish:

could you give the grass 7 daily snapshot a try?

... Replying to nkale:

That worked.

ok, good. in a few days once grass6.5 daily snapshots are fixed I might ask you to try that.

Ran with a slightly smaller dataset (didn't know if the raster format was the same between 6 & 7, so I had to do some lengthy re-importing),

for now it is the same, but that is likely to change in the future.

but I didn't get any errors, and the output map made sense.

On the other hand, in 7 I can no longer launch GRASS in MSYS mode, but that's maybe a bug for another ticket.

Known problem, being worked on.

Hamish

[hellik]

Replying to hamish:

Replying to hamish:

could you give the grass 7 daily snapshot a try?

... Replying to nkale:

That worked.

ok, good. in a few days once grass6.5 daily snapshots are fixed I might ask you to try that.

I've tried following grass65-nightly build: WinGRASS-6.5.SVN-r42351-1-Setup.exe

(Sat Jun 19 13:28:32 2010)                                                      
r.statistics base=elevation_int@user1 cover=geology_30m@PERMANENT method=sum output=elev_int_geol
(Sat Jun 19 13:28:33 2010) Befehl ausgeführt (1 sec)   

r.statistics seems to work, no error message.

Helmut

[hellik]

Replying to hellik:

Replying to hamish:

Replying to hamish:

could you give the grass 7 daily snapshot a try?

... Replying to nkale:

That worked.

ok, good. in a few days once grass6.5 daily snapshots are fixed I might ask you to try that.

I've tried following grass65-nightly build: WinGRASS-6.5.SVN-r42351-1-Setup.exe

the nightly build is from 2010/06/19, maybe revision-number is not correct? tested on WinVista32

Helmut

[martinl]

Replying to hellik:

I've tried following grass65-nightly build: WinGRASS-6.5.SVN-r42351-1-Setup.exe

the nightly build is from 2010/06/19, maybe revision-number is not correct? tested on WinVista32

right, grass65 is broken for several days. Fixed, going to recompile all branches. Martin

[hamish]

tested in 6.4 on wingrass, still fixed.