Although the docs say that you can, I can't seem to reclass a reclass:
[hnw@broke cluster-the-clusters]$ r.reclass input=IA.1000 output=IA.100 <
reclass.ameriflux1Ar1.100
ERROR: IA.1000 is a reclass of map <IA.10000> in mapset <final>. Consider
to use r.mapcalc to generate a map copy. Exiting.
[hnw@broke cluster-the-clusters]$
I'm using 5.0 pre3 ...
Just trying to save some disk space ...
Am I doing something wrong?
Bill H.
--
William W. Hargrove hnw@fire.esd.ornl.gov
Environmental Sciences Division (865) 241-2748
Oak Ridge National Laboratory (865) 574-4665 (fax)
P.O. Box 2008, M.S. 6407 http://research.esd.ornl.gov/~hnw
Oak Ridge, TN 37830-6407 FedEx/UPS, use "Bethel Valley Rd."
William W. Hargrove wrote:
Although the docs say that you can, I can't seem to reclass a reclass:
[hnw@broke cluster-the-clusters]$ r.reclass input=IA.1000 output=IA.100 <
reclass.ameriflux1Ar1.100
ERROR: IA.1000 is a reclass of map <IA.10000> in mapset <final>. Consider
to use r.mapcalc to generate a map copy. Exiting.
[hnw@broke cluster-the-clusters]$
I'm using 5.0 pre3 ...
The ability to reclass a reclass map was disabled in versions prior to
-pre4, when it was re-enabled. However, a significant bug in this
functionality was fixed in 5.0.3, so you need 5.0.3 or later (5.3/5.7)
in order to (reliably) reclass a reclass map.
In any case, there are plenty of other reasons to upgrade from -pre3
(Jan 2002), primarily the large number of bugs which have been fixed
since then.
--
Glynn Clements <glynn.clements@virgin.net>
Hi
I compiled grass 5.3 and 5.7. But make giveme an error with
gmath module (?). This error generated a lot of ignored modules.
I did make install in 5.3 and can run it, but the the ignored
modules didn't run (of course). Then i made gmake5 gmath and
giveme this error (the same in make):
[root@valinor libes]# /usr/local/bin/gmake53 gmath/
SRC = /home/chino/Bajados/Grass/grass-5.3.0/src
CMD = /home/chino/Bajados/Grass/grass-5.3.0/src/CMD
UNUSED = /home/chino/Bajados/Grass/grass-5.3.0/unused
HEADER = head.i686-pc-linux-gnu
ARCH = i686-pc-linux-gnu
GISBASE = /home/chino/Bajados/Grass/grass-5.3.0/dist.i686-pc-
linux-gnu
VERSION = 5.3.0 15 May 2004
################################################################
#
/home/chino/Bajados/Grass/grass-5.3.0/src/libes/gmath
make -f OBJ.i686-pc-linux-gnu/make.rules
gcc -I/home/chino/Bajados/Grass/grass-5.3.0/src/include -g -
O2 -c fft.c -o OBJ.i686-pc-linux-gnu/fft.offt.c: En la
función `fft':
fft.c:40: error: `fftwnd_plan' undeclared (first use in this
function)
fft.c:40: error: (Each undeclared identifier is reported only
once
fft.c:40: error: for each function it appears in.)
fft.c:40: error: error sintáctico before "plan"
fft.c:51: error: request for member `re' in something not a
structure or union
fft.c:52: error: request for member `im' in something not a
structure or union
fft.c:55: error: `plan' undeclared (first use in this function)
fft.c:58: error: `FFTW_IN_PLACE' undeclared (first use in this
function)
fft.c:66: error: request for member `re' in something not a
structure or union
fft.c:67: error: request for member `im' in something not a
structure or union
make: *** [OBJ.i686-pc-linux-gnu/fft.o] Error 1
Some help please...
Mauricio Vargas P.
Geólogo
MSc en Geología
Usurio Linux # 326558 counter.li.org
http://users.quickfox.org/~mauriciovargas/
Mauricio Vargas P. wrote:
I compiled grass 5.3 and 5.7. But make giveme an error with
gmath module (?). This error generated a lot of ignored modules.
I did make install in 5.3 and can run it, but the the ignored
modules didn't run (of course). Then i made gmake5 gmath and
giveme this error (the same in make):
[root@valinor libes]# /usr/local/bin/gmake53 gmath/
gcc -I/home/chino/Bajados/Grass/grass-5.3.0/src/include -g -
O2 -c fft.c -o OBJ.i686-pc-linux-gnu/fft.o fft.c: En la
función `fft':
fft.c:40: error: `fftwnd_plan' undeclared (first use in this
function)
The error suggests that either fftw.h or dfftw.h exists, but that it
isn't compatible. Which version of FFTW do you have? The FFT code in
the gmath library requires FFTW 2.x; it won't work with FFTW 3.x.
--
Glynn Clements <glynn.clements@virgin.net>
On Fri, 9 Jul 2004 13:49:04 +0100, Glynn Clements wrote
[...] Which version of
FFTW do you have? The FFT code in the gmath library
requires FFTW 2.x; it won't work with FFTW 3.x.
I have fftw2.x and fftw, not fftw3.
Mauricio Vargas P.
Geólogo
MSc en Geología
Usurio Linux # 326558 counter.li.org
http://users.quickfox.org/~mauriciovargas/
Mauricio Vargas P. wrote:
> [...] Which version of
> FFTW do you have? The FFT code in the gmath library
> requires FFTW 2.x; it won't work with FFTW 3.x.
I have fftw2.x and fftw, not fftw3.
Do you mean that you have both FFTW 1.x and 2.x? Having 1.x installed
may interfere with it.
Essentially, fft.c is including either fftw.h or dfftw.h, but that
file doesn't contain what it is supposed to (e.g. a definition of the
fftwnd_plan type).
--
Glynn Clements <glynn.clements@virgin.net>
On Sat, 10 Jul 2004 08:32:00 +0100, Glynn Clements wrote
Do you mean that you have both FFTW 1.x and 2.x?
Having 1.x installed may interfere with it.
Now i'm sure, i have fftw.3-x (or fftw-3.x) and fftw2.x. I
uninstalled fftw.3-x and tried to install fftw2-devel but it
need fftw2-single-devel and fftw2-double-devel; i get it now. I
will install them now and so...
Essentially, fft.c is including either fftw.h or
dfftw.h, but that file doesn't contain what it is
supposed to (e.g. a definition of the fftwnd_plan type).
Ok, thanks.
Mauricio Vargas P.
Geólogo
MSc en Geología
Usurio Linux # 326558 counter.li.org
http://users.quickfox.org/~mauriciovargas/
Hi again,
I have made a try to remove $(GMATHLIB) from the ps.map Makefile in
GRASS 5.7. Seems to work. So I have removed it in CVS as well.
Markus
Markus Neteler wrote:
I have made a try to remove $(GMATHLIB) from the ps.map Makefile in
GRASS 5.7. Seems to work. So I have removed it in CVS as well.
ps.map requires $(IMAGERYLIB), which requires G_alloc_matrix() from
$(GMATHLIB), but doesn't list $(GMATHLIB) as a dependency.
Consequently, this change causes a build failure for ps.map.
lib/imagery/Makefile should probably have EXTRA_LIBS=$(GMATHLIB) (and,
AFAICT, $(VASKLIB) as well).
--
Glynn Clements <glynn.clements@virgin.net>
On Wed, Jul 14, 2004 at 08:56:07AM +0100, Glynn Clements wrote:
Markus Neteler wrote:
> I have made a try to remove $(GMATHLIB) from the ps.map Makefile in
> GRASS 5.7. Seems to work. So I have removed it in CVS as well.
ps.map requires $(IMAGERYLIB), which requires G_alloc_matrix() from
$(GMATHLIB), but doesn't list $(GMATHLIB) as a dependency.
Consequently, this change causes a build failure for ps.map.
Ok, re-added.
lib/imagery/Makefile should probably have EXTRA_LIBS=$(GMATHLIB) (and,
AFAICT, $(VASKLIB) as well).
Isn't this already there?
Markus
On Wed, 14 Jul 2004, Markus Neteler wrote:
On Wed, Jul 14, 2004 at 08:56:07AM +0100, Glynn Clements wrote:
Markus Neteler wrote:
I have made a try to remove $(GMATHLIB) from the ps.map Makefile in
GRASS 5.7. Seems to work. So I have removed it in CVS as well.
ps.map requires $(IMAGERYLIB), which requires G_alloc_matrix() from
$(GMATHLIB), but doesn't list $(GMATHLIB) as a dependency.
Consequently, this change causes a build failure for ps.map.
Ok, re-added.
lib/imagery/Makefile should probably have EXTRA_LIBS=$(GMATHLIB) (and,
AFAICT, $(VASKLIB) as well).
Isn't this already there?
Yes I fixed it yesterday. So there is a good chance ps.map might compile OK with out the $(GMATHLIB) now. But I'm not sure.