topic Re: Assign point feature values to polygons in Geoprocessing Questions

Assign point feature values to polygons

SashikanthGurram1 — Mon, 14 May 2012 19:24:13 GMT

Hello,

I have a point feature shape file with some concentration values assigned to them. I want to assign the concentration to a 1 sq.km area with the point feature as its center. Can anyone give me some pointers in this direction?

Re: Assign point feature values to polygons

DarrenWiens2 — Mon, 14 May 2012 19:58:11 GMT

Buffer, then probably (depending on what you want to do) Identity or Union.

Re: Assign point feature values to polygons

SashikanthGurram1 — Mon, 14 May 2012 20:08:05 GMT

But, buffer yields only circular areas. What I want is a square around the point.

Re: Assign point feature values to polygons

NobbirAhmed — Mon, 14 May 2012 21:12:29 GMT

Use the buffered circles as input to Feature Envelope to Polygon tool. The output will be squares.

Re: Assign point feature values to polygons

DanLee — Mon, 14 May 2012 21:12:47 GMT

Try this:
1. Use 0.5 KM distance to produce buffers.
2. If you have ArcInfo license, use Feature Envelope To Polygon, using the buffers as input. The envelopes should be the squares you need. If you are using ArcGIS 10, you can also use Minimum Bounding Geometry tool with the geometry type, RECTANGLE_BY_AREA, to get the bounding square of the buffers. Both tools are in Data Management toolbox - Features toolset.

Re: Assign point feature values to polygons

SashikanthGurram1 — Wed, 30 May 2012 02:56:10 GMT

Thanks for the reply Dan.

The method you have suggested has solved the issue. However, I have another issue.

The point features in my dataset are tilted. However, when i create the square buffers around these points, the squares are not tilted accordingly. Thus, this leads to inconsistent square buffers. Can you suggest me something in this regard. I am attaching the image files showing the issue.'

The first image shows the tilted point features and the second image shows the created square buffers that are not tilted.

Re: Assign point feature values to polygons

DanLee — Wed, 30 May 2012 03:12:26 GMT

Oh, it seems your points are equally spaced in rows and columns; the approach would be different then. Have a look of the Create Fishnet tool:
http://help.arcgis.com/en/arcgisdesktop/10.0/help/index.html#//00170000002q000000

You need to do some work to figure out the precise extents (corner coordinates) based on the corner points plus some offsets. The resulting fishnet will be tilted with the points at centers of the cells.

Re: Assign point feature values to polygons

SashikanthGurram1 — Wed, 30 May 2012 21:13:05 GMT

It is getting difficult to estimate the outer coordinates for the vertexes of the four corner squares so as to use them as extent variables for the fishnet tool.

Is there another way where in I can rotate the squares I have attached earlier about their centroids to align them with the tilt in point features?

Alternatively, is there a way to remove the tilt from the point features, then create the square buffers and add the tilt again?

Re: Assign point feature values to polygons

DanLee — Thu, 31 May 2012 20:55:33 GMT

Ok, try the followign (still assuming points are equally spaced in both row and column directions):

1. Use 0.5 KM distance to produce buffers.
2. If you have ArcGIS 10, run Minimum Bounding Geometry tool (Data Management toolbox - Features toolset) on the buffers, specify RECTANGLE_BY_AREA for geoemtry type and ALL for Group Option. You will get one rectangle around all the buffers.
3. Zoom out so that you see a lot of while area around your points. Run Create Thiessen Polygons tool (Analysis toolbox - Proximity toolset; requiring ArcInfo license) on the points; make sure to set Environment ... - Processing Extents to be "Same as Display". You should get square polygons around points, except at the edges where the polygons fill up the display.
4. Use Clip tool (Analysis toolbox - Extract toolset) to clip the Thiessen polygons by the rectangle from Step 2 above.

Is the result what you need?

If you want to automate the "Zoom out" part of Step 3, you can just set a large extent values, so that the Thiessen polygons go out of the rectangle.

Re: Assign point feature values to polygons

SashikanthGurram1 — Fri, 01 Jun 2012 19:26:27 GMT

Thanks Dan. This brought me to closest point I want to be. Please have a look at the first image attached. The polygons have tilted along with the points which is what I want.

When I zoomed in, I found that the shapes created using theissen polygons have additional sides at two vertices thus making it a hexagon. Please look at the second picture attached.

Can we do anything about this?

Re: Assign point feature values to polygons

DanLee — Fri, 01 Jun 2012 20:35:42 GMT

Have you done the Step 4: Clip? I wonder if you can set the XY tolerance in Clip so that the two closely located vertices snap together, if the default xy tolerance is not enough.

Otherwise:
- Run Integrate tool (Data Management toolbox - Feature Class toolset) with the XY tolerance large enough to get the two closely located vertices snap together. The result shapes may not be perfect squares. How critical is the shape to your analysis?

Re: Assign point feature values to polygons

SashikanthGurram1 — Sun, 03 Jun 2012 04:54:43 GMT

Have you done the Step 4: Clip? I wonder if you can set the XY tolerance in Clip so that the two closely located vertices snap together, if the default xy tolerance is not enough.

I have done the step 4 which yielded the polygons with a line at the vertices. So, I have increased the tolerance and now what I got is almost like a square around the points. Certainly, your suggestion has helped me.

Otherwise:
- Run Integrate tool (Data Management toolbox - Feature Class toolset) with the XY tolerance large enough to get the two closely located vertices snap together. The result shapes may not be perfect squares. How critical is the shape to your analysis?

I haven't done this step as increasing the tolerance while clipping solved the issue. I have calculated the area for each grid around the points and it is almost equal to what my expectation is. I believe there are some discrepancies with regard to the point layout which led to this. However, these discrepancies appear to be minor.

Thanks for all the help.