Determining Fixed Distance Band for Hot Spot Anallysis

4566
1
04-29-2010 09:04 AM
AlanWu
by
New Contributor
I doing some hot analysis on distribution of restaurants in LA County. I am rather new to this and need some help in determining the distance band for hot spot analysis. I have a LA County census polygon feature layer with the number of restaurants per census tract. I am trying to determine the distance band to run the hot spot analysis to determine if there is any clustering pattern of food sites in LA County.

When I tried to run the Moran I to determine the distance band without entering any value and use the default distance band it gives me a value of .2265. I am not sure what units this is in whether it's meter or feet or whether this makes any sense or whether I am doing this right.

How do I make sure that the distance values are in the same units as the output coordinate system?

Is it okay to use not enter any value and used the default which comes out to .2265?

I tried distance bands of 5, 10, 50, 1000 and I get a warning and error message:

WARNING: The input feature class does not appear to contain projected data.

ERROR: The threshold distance must be smaller than the feature class extent.

Would greatly appreciate some help and guidance.
Tags (2)
0 Kudos
1 Reply
DanPatterson_Retired
MVP Emeritus
your coordinates are in decimal degrees, aka Geographic, and need to be projected to a coordinate system that uses feet or meters (ie UTM etc), use the Project tool in Arctoolbox ( http://help.arcgis.com/en/arcgisdesktop/10.0/help/index.html#//00170000007m000000.htm ) or project the data in ArcCatalog.  The old forums contain hundreds of posts related to projecting data
http://forums.esri.com/forums.asp?c=93
and equivalent help is in the older version help files ( http://webhelp.esri.com/arcgisdesktop/9.3/index.cfm?id=1890&pid=1887&topicname=Project_(Data_Managem...) )
0 Kudos