Percentage hotspot/density

[JayneEvans]

Hello all,

I have some property points for a city. In the data I have a field with either a 0 or 1 dependent on the type of property each point is. I would like to be able to map the percentage of properties which are type 1 across the whole city so we can find hotspots of the specific property type. We are particularly interested in being able to look at different percentages so that we can locate areas where, for example, the percentage of properties that are type 1 is above 20% and below 80%. It would be perfect if we could visualise the percentages as a surface (so maybe we could symbolise by percent band i.e. 0-20%, 20-50%, 50-80%, 80-100%).

I've been researching the different density/hotspot tools and tying myself in knots as to what would be an appropriate technique to use. I have tried point density and hotspots but am a bit confused by the figures it outputs (stats are not my strong point!).

Any advice/suggestions would be very much appreciated. If I having explained anything fully please ask.

Thanks,
Jayne

[curtvprice]

I have some property points for a city. In the data I have a field with either a 0 or 1 dependent on the type of property each point is. I would like to be able to map the percentage of properties which are type 1 across the whole city so we can find hotspots of the specific property type. We are particularly interested in being able to look at different percentages so that we can locate areas where, for example, the percentage of properties that are type 1 is above 20% and below 80%. It would be perfect if we could visualise the percentages as a surface (so maybe we could symbolise by percent band i.e. 0-20%, 20-50%, 50-80%, 80-100%).

Given decision makers will be implementing decisions by political area (ward, precinct, etc) you may want to do this with feature data, ie do an Identity (or join by location) of the your property centroid points with political unit polygons and then summarize numbers of "type 1" and "other" points in each polygon. (A nice trick to this is to assign them scores of 0 and 100 and calculate a mean - this computes to percent.)

This approach may also avoid problems with your spatial distribution of points skewing your results.

I realize this isn't a Spatial Analyst solution to your problem but it may be a better place to start. You can get fancier with hot spot analysis once you know the basic patterns.