1

I have several hundred groups of data points (in lat/lng format). I want to generate heat maps for each group using kriging or similar interpolations and then display the generated raster images on a map server. What will be the best solution? I can't use ESRI products as they're not free. I know a little bit of Python.

Thanks!

underdark
  • 84,148
  • 21
  • 231
  • 413
sean hawk
  • 119
  • 1

3 Answers3

3

Google's Fusion Tables is a free product, and has the ability to make heat maps from points.

You would load your lat/long points into the fusion table, specify that the fields denoted lat/long, and check the Heatmap option. There's a good example here.

Note that there's no control over how the heatmap is generated, so it may not suit your advanced purposes given that you mentioned kriging, Python, etc.

Stephen Lead
  • 21,119
  • 17
  • 113
  • 240
  • I kind of remember the heatmap generated by google fusion is by point density; that is, no z-value is used. I hope there is a kriging/interpolation library that I call using Python. In addition, I don't know what is a good way to store the images and display them. I hope there's a framework that is easy to configure. I'm kinda new in GIS. Thanks! – sean hawk Dec 07 '11 at 00:31
  • http://gis.stackexchange.com/questions/1041/open-source-methods-for-kriging might help – Stephen Lead Dec 07 '11 at 00:41
0

You can try this Python program here. It gives some pretty heatmap visualizations. But the output is not geo-referenced. You can them as images are even videos.

Arunmozhi
  • 36
  • 2
0

I'd go with a combination of tools. R sounds like the best fit for your interpolations and raster formatting. As far as web display is concerned, you have options that include a PostGIS/postGreSQL stack for a 'true' web mapping service. D3.js has good mapping tools (including tile support) that may be appropriate if you're trying to just get something out there fast.

metasequoia
  • 559
  • 5
  • 17