Hustota bodů a geokódovaná data

LarryRuiz

<\/HEAD>

Snažím se použít nástroj Point Density k určení hustoty pacientů v naší síti. Vzal jsem dávku dat pacientů a geokódoval je pomocí nástroje locator ze Streetmap Premium. Vzal jsem geokódovanou vrstvu a použil nástroj Point Density, ale dostávám pouze jeden rozsah. Mám pocit, že mi chybí nějaký krok. Máte nějaké návrhy?<\/P><\/BODY><\/HTML>

DanPatterson_Retired

Well you will have to move beyond defaults and think about what it is contained in your data based upon its geometric spread within your study area, the attributes in your table..  for example...

Field denoting population values for each point. The population field is the count or quantity to be used in the calculation of a continuous surface.

Values in the population field may be integer or floating point.

The options and default behaviors for the field are listed below.

• Use None if no item or special value will be used and each feature will be counted once.
• You can use Shape if input features contains Z.
• Otherwise, the default field is POPULATION. The following conditions may also apply..... etc etc etc

​I presume that you used the default of None meaning that your points aren't weighted with some attribute and will only be counted once.  Which means, that every location will be treated equally and there is nothing special about a point other than its location. 

Now on to cell size...

The cell size for the output raster dataset.

This is the value in the environment if specifically set. If the environment is not set, then cell size is the shorter of the width or height of the output extent in the output spatial reference, divided by 250.

Most people ignore this when they first encounter a tool, but it has implications.  If you let 'it' do its thing, you can get unforseen results.  If your study area has points that are relatively spread out, then the default might be appropriate, but consider the following point pattern (first little bit anyway)

....      . .                 . . .                              .......

....                .                             .              .......

do you get the impression that there is a grouping to the left and a grouping to the right?  Well if you don't, read no further.  If you do...imagine what is going to happen when you let it divide your extent by 250 to produce cells...most of the information in the point grouping to the left and right is going to get agglomerated into a couple (at best) cells whereas the presence of those few measly points in the middle get over-emphasized relative to the rest of the points.  Your 'pattern' will be different depending upon the cell size.

I will refrain from going on about the affect that the choice of neighbourhood will have and the output units etc etc, as well as the other options within the toolset to An overview of the Density toolset—Help | ArcGIS for Desktop  which may also be appropriate.   This is one of those questions, where I should have said "what do your points look like" .... an image would help and perhaps 'Someone' could comment on the results given their suggestion.

Keep us posted