Latest Contributions by OlofOlsson

‎11-20-2015

I am using Summary statistics to calculate a new attribute based on the previously calculated Match attribute, gropued by the polygon id's (there are multiple points with the same id). This allows me to select the closest match and is the best solution I can think of right now. Thank you for your help. Edit: rather than to extract the MIN or MAX values, I used the square root of the Match attribute since the best match often lies somewhere between the Min and Max values. Seems to work for my material anyway

‎11-20-2015

Well right now I'm doing it manually. I'm running Identity to connect the points with the polygons and then calculate their respective match as a percentage between the area attributes. Since I will do this for quite many features I am really looking for a more automatic way...

‎11-20-2015

Hi! I need to join polygons with points. For each shapefile I have attributes that represent the same information (area), but the problem is that these attributes do not perfectly match. Since I have multiple points per polygon, I don't want to do a spatial join, and the distance from the point to the polygon is not relevant here. So my question is if there is any way to join points with polygons by their closest matching values (area)? Regards,

‎11-06-2013

Hi! Is it possible to run a single GWR on separated regions (e.g. city blocks in 10 different cities across a county), or does it defeat the purpose of GWR?

‎01-29-2013

Hi, I want to run a geographically weighted regression on a polygon feature class with missing values (-999) for some of the cases. Is there a way to convert missing values to NoData in a polygon? I have used SetNull before but I think it only works on raster data. If it works to first convert the data to raster, use SetNull, and then convert it back again to polygons and use it in a gwr I could off course do that, but I dont know how these missing values would be identified in a polygon... /Olle

‎01-23-2013

Thanks, I will look into that. I am currently using submodels (building my main model for the closest facility analysis and then adding that 47(!) times in a new model, then manually specifying the incidents and facilities for each city before I run the model). It never occured to me to use iterators like you suggested. This is off course a much more elegant solution 🙂 You are right, a model can contain only one iterator. How you can create a model which will contain a model (just drag and drop the model from Catalog window to the canva) and add another iterator if needed.

‎01-23-2013

Hi Olle, This scenario is possible to implement by using the Add Locations or Make Closest Facility Layer GP tools (depending on your needs) and the Iterate Feature Selection iterator in ModelBuilder. Check how the iterator works here. Thank you for your quick reply. I have already built the model in model builder but since I have two feature classes for facilities and origins I dont think I can use an interator to solve this (since model builder only supports one iterator per model). But maybe I didnt understand you correctly? /Olle

‎01-23-2013

Hi, I have to run Closest facility for 47 cities separately. I have a model for this but dont want to run it manually for each city. Is there a way, for example by using Select or Select by attribute, to add the facilities and incidents for each city in a single model in model builder? /Olle

‎01-22-2013

Hi, I have a strange problem with the drivetime attribute from an OD Cost Matrix in ArcGIS 10. The analysis solves without problems, but the calculated drivetime attribute in Routes seems to be based on the length of the lines used to symbolize the routes, not the length of the streets. If I recalculate the length of the street segments (in the Routes layer), I can also recalculate the drivetime. However, I need to do my analysis for 47 cities separately in model builder and want to avoid manual work.... I have used a sample of my origin and destination points with the Closest facility and the Service area (with lines) solvers to check for errors on the network but they solve the routes just fine with correct drivetimes. Any ideas on how to fix the drivetime calculation in routes? Perhaps it is a performance issue? I have 741 destinations and 280 origins. When I only solve the routes between the origins and destinations that are causing the problem it works fine, with correct drivetimes... Edit: For anyone interested, I identified the problem as related to the origins not being properly calculated by the Calculate locations tool. It seems to work now. /Olle

‎12-05-2012

Hi! I am doing a fuzzy overlay analysis and was wondering if I can plot the relationship between the observed (continuous) and the fuzzy membership values in ArcGIS 10? Perhaps I need to combine the values before the plot somehow? Edit: I have tried to use Combine but this transforms the Membership values into integers. Edit 2: I got it to work by using Int("Variable" * 1000) on each of the two grids before running Combine, and then dividing each variable by 1000 to extract their original values in the attribute table of the Combine raster. I found the solution in another post (http://forums.arcgis.com/threads/35263-preserving-original-raster-values-when-combining). Hence this thread may now be locked, if there are no alternative or more straightforward solutions to my problem. /Olle

‎05-31-2012

Hi, You could do this by using the Near tool. This will at least give you the Euclidean distance between each point in your dataset. Hope this helps! Olle

‎03-14-2012

Hi! I want to extract the metadata from one polyline onto another polyline without creating a new layer. However, the polylines (two road layers) contain different levels of generalization and do not always overlap (see attachment). I figured I could save some time by letting Arcmap 10 do most of the job, and then I could edit the rest manually (eg. information about speed limits).

‎09-08-2011

Hi! I am examining land cover changes in 1990-2010 based on Landsat 7 satellite data, which I classify to raster images using the image classification toolbar in Arcmap 10. Is there a way to compare the classification quality, or ascertain that an equal quality level has been reached, between different raster images? I know of the sample evaluation tools, and one can always visually compare the results with satellite or aerial photos, but is there a more precise way? Olle

Online Status	Offline
Date Last Visited	‎11-11-2020 02:24 AM

My Ideas

Latest Contributions by OlofOlsson

Re: Join to closest matching attribute

Re: Join to closest matching attribute

Join to closest matching attribute

GWR on separated regions

GWR on feature class with missing data

Re: Add multiple locations in Model builder

Re: Add multiple locations in Model builder

Add multiple locations in Model builder

Drivetime in OD Cost Matrix

Plot observed and fuzzy membership values

Re: Distance matrix

Combine metadata from non-overlapping polylines

How to compare quality between two classified images?