[Qgis-user] problems making grass training map via qgis

Mon Aug 27 00:02:22 PDT 2012

On 2012/08/24 21:29, Micha Silver wrote:
> Hello Sam:
>
> On 24/08/2012 14:14, Samantha Goodchild-Brown wrote:
>> Mischa
>>
>> I am working on those commands but In Linux side I have got the 
>> training map to raster format with all visible.
>> More issues I overcame (for students benefit) is:
>> I imported a shape file made in Qgis with the RGB column and label 
>> column. To get this to work I had to select 'command dialog' button 
>> in 'import vector data' box and not just use' ivd' box gui.
>> In the' import vector data' gui I ticked 'override dataset projection 
>> (use locations projection)' and 'limit import to the currect region' 
>> and made no other changes in the 
>
> That's unusual that you had to specify "override dataset projection". 
> Are you sure that the original is projected in the same coordinate 
> system as the GRASS LOCATION??
Yes I am sure. However the Raster is a new geotiff and the shape file is 
old style. Evidently Grass doesn't recognise them as the same. When you 
read the metadata in QGIS you can understand why but it is still quite 
frustrating.
>> command dialog box except to repeat input and output dataset names!
>> Now v.to rast works but it doesn't allow out to specify the attribute 
>> columns - looks greyed out, so you just type them as you know them, 
>> it works!
>>
>> In windows grass I imported the same shp file, in the same way, 
>> successfully. (this using grass not grass plugin in QGIS)  In the 
>> v.to.rast I can specify / select the attribute columns but when I 
>> 'run' it, I get the following:
>>
>> (Fri Aug 24 12:46:44 2012)
>> v.to.rast input=qgis_lclu_trainingmap2 at 3420new 
>> output=trainingmap_raster column=cat_ rgbcolumn=RGB labelcolumn=label
>> Loading data...
>> Reading areas...
>> Reading features...
>> Writing raster map...
>> Converted areas: 6 of 6
>> Converted points/lines: 0 of 0
>> v.to.rast complete.
>> (Fri Aug 24 12:46:45 2012) Command finished (0 sec)
>>
>> Which one would think means everything is fine but now can see no 
>> data - the entire dataset created is the same colour (white).
>>
>> Is the data most likely still there and just not displaying 
>> correctly, or has some of the attribute info all been lost and it all 
>> does have the same value?
>
> Did you set the current region to match the vector or raster of 
> training areas?  ("Settings->Region->Set Region" and choose the 
> imported raster). Then in the map display use the zoom button to "Zoom 
> to computational region".
Yup. Ah, no wait. There is a problem here.
OK, now I see. Set the computational region to be the import raster 
before working. I didn't see that the project region is different from 
the computational region.
I definitely had g.region set but 'Then in the map display use the zoom 
button to "Zoom to computational region"' I had completely missed.
I think it will always be a good idea to have 'show comp. extent' ticked on!
Many thanks.
>
> You can also try "Vector->Query vector map->Query attribute data" to 
> see what you've got as far as attributes, and "Vector->Reports and 
> statistics->Basic metadata" to make sure that you have as many 
> polygons as you expect.
>
>
>>
>> On 2012/08/16 18:48, Micha Silver wrote:
>>> On 16/08/2012 18:11, Samantha Goodchild-Brown wrote:
>>>> Thank you
>>>>
>>>> In my frustration I have reverted to linux version which has proved 
>>>> useful - seeing more messages is helpful.
>>>> Mischa, your tip of g.region vect as well as finding reference to 
>>>> v.in.region has proved helpful. This is likely to be my critical 
>>>> mistake.
>>>> In using the 'grass interface' I have again hit the RRR:GGG:BBB 
>>>> field request (as I did with windows).
>>>> It is very frustrating to find this now again when I can find no 
>>>> other reference to it anywhere (not even in tuts cautioning to make 
>>>> label column).
>>>> Is it absolutely neccessary to go into this at this point (go all 
>>>> the way back and begin again) or is there a way to 'bypass' it?
>>>> Or some quick shortcut?
>>>>
>>>
>>> Can you copy into an email the commands you are running, and the 
>>> results (errors) you are getting?
>>> If you're using the Linux terminal, then copy/paste the commands and 
>>> output, and post here.
>>>
>>>>
>>>>
>>>> Sam
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> On 2012/08/15 19:52, Paolo Cavallini wrote:
>>>>> Or use OTB, also through sextante.
>>>>> --
>>>>> http://faunalia.eu/pc
>>>>> Sent from mobile, sorry for being short
>>>>>
>>>>> richard burcher <drownedfrog at gmail.com> ha scritto:
>>>>>
>>>>>     Another option is to create training polygons&  use [1] SAGA for the
>>>>>     supervised classification (either standalone or through the Sextante
>>>>>     plugin). I've recently gone this way for a maximum likelihood
>>>>>     classification.
>>>>>
>>>>>     [1]http://www.saga-gis.org/en/index.html
>>>>>
>>>>>     Cheers,
>>>>>
>>>>>     Richard
>>>>>
>>>>>
>>>>>
>>>>>     --
>>>>>     View this message in context:http://osgeo-org.1560.n6.nabble.com/problems-making-grass-training-map-via-qgis-tp4995324p4995387.html
>>>>>     Sent from the Quantum GIS - User mailing list archive atNabble.com  <http://Nabble.com>.
>>>>>     ------------------------------------------------------------------------
>>>>>
>>>>>     Qgis-user mailing list
>>>>>     Qgis-user at lists.osgeo.org
>>>>>     http://lists.osgeo.org/mailman/listinfo/qgis-user
>>>>>
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> Qgis-user mailing list
>>>>> Qgis-user at lists.osgeo.org
>>>>> http://lists.osgeo.org/mailman/listinfo/qgis-user
>>>>
>>>>
>>>>
>>>> This mail was received via Mail-SeCure System.
>>>>
>>>>
>>>> _______________________________________________
>>>> Qgis-user mailing list
>>>> Qgis-user at lists.osgeo.org
>>>> http://lists.osgeo.org/mailman/listinfo/qgis-user
>>>>
>>>> This mail was received via Mail-SeCure System.
>>>>
>>>>
>>>
>>
>>
>>
>> This mail was received via Mail-SeCure System.
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/qgis-user/attachments/20120827/e80c4da0/attachment.html>