# How to create zip code areas from point data?

I have a point dataset which has zipcodes for a certain region. This zipcodes have been marked on the basis of lat long collected. So each zipcode is repeated as it would be useful to show the region covered by one zipcode. Now since there are points I would like to know if there is any tool in QGIS or ArcGIS which can make polygons for each zipcode.

-
Do you have one point per zip code? If you have multiple points per zip code, are they corner points of the zip code areas or rather some kind of center points? – underdark Feb 13 '12 at 7:53
These are random points . One zipcode may have a few points and a certain zipcode may have just one point also. – SNT Feb 13 '12 at 8:20
This appears to be substantially the same as gis.stackexchange.com/questions/20020/points-to-polygon and has exactly the same answers. – whuber Feb 13 '12 at 17:14

In Vector - Geometry tools you'll find the Voronoi Polygons tool. It will create an area around each point. Afterwards, you can merge areas with same zip code.

This will only give a rough approximation of the real shape of the zip code areas but I don't see how to get any closer to the real shape with the input data you have.

-
I was thinking of converting this point data into raster on the basis of the pincode as the Z value but somewhat unable to do it here as it just gives me a single band raster output. Inital idea was to convert them to raster and again raster to polyon .Makes sense ?? – SNT Feb 13 '12 at 10:18
No, I wouldn't convert the points to raster. – underdark Feb 13 '12 at 10:31
Created Voronoi polygons with a buffer of 10%. Then dissolved all the polygons with same pincode values. And it gives a much better result so far. – SNT Feb 13 '12 at 11:23
I did finally get the voronoi polygons.I wanted to know if there is a way where I could do a proximity analysis which would give me an output in any form . The idea is say for one instance a polyon which covers a certain pin code I would like to get an analysis done on this data which would give all the surrounding pincode polygons to one pincode polygon. – SNT Feb 13 '12 at 12:53

The Voronoi diagrammer is the best fit of this polygon representation, but the quality of the results will be related to the quality of your point data.

If you have the polilynes of the zip code, maybe transforming them into point data could make your results better.

After the Voronoi, don't forget to do the dissolve using the Zipcode.

-