Hello.
I am making a variable selection from Landsat(6 optical spectral bands),
ndvi, tasseled cap and Texture for every band to apply classification
algorithms.
I would like to select the best variables based on the best criterion. I
found in [http://www.stat.umn.edu/geyer/5102/examp/select.htm
Hi,
yes, the quotes are indeed tricky... The challenge is to provide the
correct mix of escaped and un-escaped single and double quotes to Python
/ ArcGIS as in this Python example:
gp.select_analysis("nfroads.shp", "paved.shp",
' "ROAD_CLASS" = \'PAVED\' ')
In order to achieve this, the
Hi,
I like to compare areas not exceeding of particular threshold values
between parametric and non-parametric approaches of 1000 realizations.
For non-parametric approach, I did IK for 1000 realizations (conditional
simulation) and calculated the area not exceeding (0.5 probability) of
the
Dear Arthur,
I have not seen any responses, therefore I will have a go.
I would start to interpolate the continuous risk score using kriging
(e.g. krige in the gstat package). This will give you a continuous map.
Then you could categorise that map. Kriging requires that you have only
one value pe
Andrew, thanks for your reply!
I tried exactly your command, but still the resulting rpygeo.py file
contains the following line:
gp.select_analysis( "nfroads.shp", "paved.shp", "\"ROAD_CLASS\" = 'PAVED' "
)
So the \"ROAD_CLASS\" = 'PAVED' is still surrounded by double quotes ("). I
have tried ma
Maarten
I'm not sure about the single quotes but possibly you could add an extra
special character break (the \ ) to indicate to python that the double quotes
surrounding your field name are not the end of a string. You could try the
following code:
rpygeo.geoprocessor("select_analysis", c(
Many thanks to everyone !
On Thu, Dec 17, 2009 at 8:47 PM, Roger Bivand wrote:
> On Thu, 17 Dec 2009, Dan Putler wrote:
>
>> Hi all,
>>
>> For this particular problem, Dylan Beaudette is probably exactly right.
>> However, I ran into problems using similar tools in OpenJump when we
>> wanted to c
