[R-sig-Geo] rgrass7 : Error in parseGRASS
Chris Reudenbach
reudenbach at uni-marburg.de
Wed Nov 23 09:58:50 CET 2016
Michael,
using GRASS from R is a bit cumbersome. GRASS (and this is for sure one
of it strength ) is extremely rigid with projections etc. Due to this
you should setup a correct structure. I use a setup that is similar to
the below snippet.
cheers Chris
library(rgrass7)
library(sp)
# setup a temp workingdir
working.dir<- "~/tmp/"
# get meuse data
data(meuse)
data(meuse.grid)
# georeference the meuse grid data
coordinates(meuse.grid) <- ~x+y
proj4string(meuse.grid) <- CRS("+init=epsg:28992")
gridded(meuse.grid) <- TRUE
# get a first cellsize/pixel resolution for GRASS
resolution <- meuse.grid at grid@cellsize[1]
# georeference the meuse data
coordinates(meuse) <- ~x+y
proj4string(meuse) <- as.character(CRS("+init=epsg:28992"))
# get projection, proj4 string and extent for GRASS
projection<-(strsplit(meuse at proj4string@projargs,split = " "))
proj4<- paste(projection[[1]][2:length(unlist(projection))], collapse = ' ')
xmax<-meuse at bbox[3]
xmin<-meuse at bbox[1]
ymax<-meuse at bbox[4]
ymin<-meuse@@bbox[2]
# create and set working directory
if (!file.exists(file.path(working.dir,"run"))){
dir.create(file.path(working.dir,"run"),recursive = TRUE)
}
setwd(file.path( working.dir,"run"))
############ set up GRASS RUNTIME environment
# define the GRASS executable path
if(Sys.info()["sysname"] == "Windows"){
grass.gis.base<-'C:/OSGeo4W64/apps/grass/grass-7.0.5'
}else {
grass.gis.base<-'/usr/lib/grass72'
}
# set path for optional GRASS addons
Sys.setenv(GRASS_ADDON_PATH="~/.grass7/addons")
# create the TEMPORARY GRASS location
rgrass7::initGRASS(gisBase=grass.gis.base,
home=tempdir(),
mapset='PERMANENT',
override=TRUE
)
# assign GRASS projection according to data set
rgrass7::execGRASS('g.proj',
flags=c('c','quiet'),
proj4=proj4
)
# assign GRASS extent and resolution
rgrass7::execGRASS('g.region',
flags=c('quiet'),
n=as.character(ymax),
s=as.character(ymin),
e=as.character(xmax),
w=as.character(xmin),
res=as.character(resolution)
)
############# now do GRASS STUFF
writeVECT(meuse,"meuse")
execGRASS("v.db.addcolumn", map = "meuse", columns = "area_m2 double
precision")
On 23.11.2016 07:38, Michael DELORME wrote:
> Thanks for your insight.
> I'll use a stand-alone GRASS install.
>
> Cordially
>
> Le 22/11/2016 15:13, Roger Bivand a écrit :
>> On Tue, 22 Nov 2016, Michael DELORME wrote:
>>
>>> Thanks for investigating.
>>>
>>> I'll try a standalone version of GRASS if needed.
>>>
>>> Here is the traceback
>>>
>>>> traceback()
>>> 4: stop(paste(cmd, "not parsed"))
>>> 3: parseGRASS(cmd, legacyExec = legacyExec)
>>> 2: doGRASS(cmd, flags = flags, ..., parameters = parameters, echoCmd =
>>> echoCmd, legacyExec = legacyExec)
>>> 1: execGRASS("v.db.addcolumn", map = "meuse", columns = "area_m2 double
>>> precision")
>>>
>>> My C:\OSGeo4W64\apps\grass\grass-7.0.5\bin lists *.exe or *.bat and
>>> indeed the *.bat command are those which don't work (r.mask.bat,
>>> v.db.addcolumn.bat,...). The batch calls Python scripts ; for example
>>> v.db.addcolumn.bat is :
>>>
>>> @"%GRASS_PYTHON%" "%GISBASE%/scripts/v.db.addcolumn.py" %*
>>>
>>> So I guess something must be wrong in my Python installation... (it
>>> works however from within a GRASS shell)
>> No, the logic for Windows and initGRASS has only ever been tested with
>> stand-alone GRASS. I would guess that in OSGeo4W the path variables
>> are wrong. I cannot even find out how to run initGRASS in R but
>> outside the OSGeo4W shell (doesn't find iconv.dll) or inside the
>> OSGeo4W shell (cannot create a temporary GRASS location). Unless you
>> or others really need OSGeo4W, I suggest asking others to implement
>> that - very bulky and fragile.
>>
>> Roger
>>
>>> Cordially
>>>
>>>
>>> Le 22/11/2016 14:30, Roger Bivand a écrit :
>>>> On Tue, 22 Nov 2016, Michael DELORME wrote:
>>>>
>>>>> Thanks for replying
>>>>>
>>>>> Here is a reproducible example (change your GRASS directory of
>>>>> course) :
>>>>>
>>>>> library(rgrass7)
>>>>> library(sp)
>>>>> initGRASS("C:/OSGeo4W64/apps/grass/grass-7.0.5", home = tempdir(),
>>>>> override = TRUE)
>>>>> data(meuse)
>>>>> coordinates(meuse) <- ~x+y
>>>>> writeVECT(meuse, "meuse")
>>>>> execGRASS("v.db.addcolumn", map = "meuse", columns = "area_m2 double
>>>>> precision")
>>>> This appears to run correctly on Windows 7 with GRASS 7.0.5 Windows
>>>> stand-alone. Do you need to use OSGeo4W - the internal logic used for
>>>> finding out where the different versions keep their executables and
>>>> batch files varies?
>>>>
>>>> What does traceback() say after the error? The error comes from
>>>> parseGRASS("v.db.addcolumn") - the XML error comes from not being able
>>>> to read the output of v.db.addcolumn<.ext> --interface-description
>>>> where <.ext> may be .exe or .bat I think.
>>>>
>>>> Roger
>>>>
>>>>> I get :
>>>>> Error : XML content does not seem to be XML: 'The specified path
>>>>> was not
>>>>> found.'
>>>>> In addition: Warning message:
>>>>> running command 'v.db.addcolumn.bat --interface-description' had
>>>>> status 1
>>>>> Error in parseGRASS(cmd, legacyExec = legacyExec) :
>>>>> v.db.addcolumn not parsed
>>>>>
>>>>> Other execGRASS commands (like r.neighbors, r.mapcalc,..) work fine,
>>>>> but some other (r.mask) fail with a similar XML error.
>>>>>
>>>>> All commands work fine when used from the GRASS shell...
>>>>>
>>>>> What is this XML error ?
>>>>> Thanks
>>>>>
>>>>> Le 22/11/2016 09:34, Roger Bivand a écrit :
>>>>>> On Tue, 22 Nov 2016, Michael DELORME wrote:
>>>>>>
>>>>>>> Dear list,
>>>>>>>
>>>>>>> Using rgrass7 on Windows7 with GRASS 7.0.5 (OSGeo distribution),
>>>>>>> I get
>>>>>>> the following error :
>>>>>> Please always try to provide a reproducible example using a standard
>>>>>> data set. In addition, the output of traceback() following the error
>>>>>> would be helpful. Are you running R inside GRASS, or using R to
>>>>>> initiate GRASS?
>>>>>>
>>>>>> Consider using the grass-stats list:
>>>>>>
>>>>>> https://lists.osgeo.org/mailman/listinfo/grass-stats
>>>>>>
>>>>>> as it is more likely to yield results, but there also provide an
>>>>>> example that anyone can reproduce. The example should avoid loading
>>>>>> other packages, and should be run from the R console (not, say,
>>>>>> RStudio).
>>>>>>
>>>>>> Roger
>>>>>>
>>>>>>
>>>>>>
>>>>>>>> execGRASS("v.db.addcolumn", map = "result_v", columns = "area_m2
>>>>>>>> double precision")
>>>>>>> Error : XML content does not seem to be XML: 'Le chemin d'accés
>>>>>>> spécifié est introuvable.' In addition: Warning message: running
>>>>>>> command 'v.db.addcolumn.bat --interface-description' had status 1
>>>>>>> Error in parseGRASS(cmd, legacyExec = legacyExec) :
>>>>>>> v.db.addcolumn not parsed
>>>>>>>
>>>>>>>
>>>>>>> and a similar one for r.mask :
>>>>>>>
>>>>>>>> execGRASS("r.mask", raster = "map")
>>>>>>> Error : XML content does not seem to be XML: 'Le chemin d'accés
>>>>>>> spécifié est introuvable.'
>>>>>>> In addition: Warning message:
>>>>>>> running command 'r.mask.bat --interface-description' had status 1
>>>>>>> Error in parseGRASS(cmd, legacyExec = legacyExec) : r.mask not
>>>>>>> parsed
>>>>>>>
>>>>>>>
>>>>>>> whereas most other execGRASS command I tried work as expected.
>>>>>>> Any idea ?
>>>>>>> Thanks in advance
>>>>>>>
>>>>>>> Here is my session info :
>>>>>>>
>>>>>>>> sessionInfo()
>>>>>>> R version 3.3.1 (2016-06-21)
>>>>>>> Platform: x86_64-w64-mingw32/x64 (64-bit)
>>>>>>> Running under: Windows 7 x64 (build 7601) Service Pack 1
>>>>>>>
>>>>>>> locale:
>>>>>>> [1] LC_COLLATE=French_France.1252 LC_CTYPE=French_France.1252
>>>>>>> LC_MONETARY=French_France.1252 LC_NUMERIC=C [5]
>>>>>>> LC_TIME=French_France.1252
>>>>>>> attached base packages:
>>>>>>> [1] parallel stats graphics grDevices utils datasets
>>>>>>> methods base
>>>>>>> other attached packages: [1] rgrass7_0.1-9 XML_3.98-1.4 beepr_1.2
>>>>>>> raster_2.5-8 sp_1.2-3 RPostgreSQL_0.4-1 DBI_0.5-1
>>>>>>>
>>>>>>> loaded via a namespace (and not attached): [1] magrittr_1.5
>>>>>>> tools_3.3.1 audio_0.1-5 Rcpp_0.12.7 stringi_1.1.2 grid_3.3.1
>>>>>>> stringr_1.1.0 lattice_0.20-34 [9] fortunes_1.5-3
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> [[alternative HTML version deleted]]
>>>>>>>
>>>>>>> _______________________________________________
>>>>>>> R-sig-Geo mailing list
>>>>>>> R-sig-Geo at r-project.org
>>>>>>> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>>>>> --
>>>>> *Michaël DELORME*
>>>>> Administrateur de données, chargé de projet en géomatique
>>>>> Service régional de l'information statistique, économique et
>>>>> territoriale
>>>>> Direction régionale de l'alimentation, de l'agriculture et de la forêt
>>>>> Auvergne-Rhône-Alpes
>>>>> Site de Lyon
>>>>> 165 rue Garibaldi – BP 3202 – 69401 LYON cedex 03
>>>>> 04 78 63 34 21
>>>>>
>>>>> Pour écrire au service :
>>>>> DRAAF Auvergne-Rhône-Alpes – SRISET
>>>>> 16B rue Aimé Rudel – BP 45 – 63370 LEMPDES
>>>>>
>>>>> http://www.draaf.auvergne-rhone-alpes.agriculture.gouv.fr/
>>>>>
>>>>> [[alternative HTML version deleted]]
>>>>>
>>>>> _______________________________________________
>>>>> R-sig-Geo mailing list
>>>>> R-sig-Geo at r-project.org
>>>>> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>>>
>>>
> _______________________________________________
> R-sig-Geo mailing list
> R-sig-Geo at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>
More information about the R-sig-Geo
mailing list