Message Boards Message Boards

0
|
11342 Views
|
4 Replies
|
3 Total Likes
View groups...
Share
Share this post:

Difficulty with PointProcess

When I use PoissonPointProcess with my data, it is OK

ClearAll["Global`*"]
villes = Entity["Country", "France"]["LargestCities"];
data = SpatialPointData[villes]
proc = EstimatedPointProcess[data, PoissonPointProcess[\[Lambda], 2]]
ListPlot[{RipleyK[data, Range[0, 200, 10]], 
  RipleyK[proc, Range[0, 200, 10]]}, 
 PlotLegends -> {"original process", "estimated model"}]

But if I work with an other PointProcess, I can't have a result. Why ?

see for example :

proc4 = EstimatedPointProcess[data, StraussPointProcess[a, b, c, 2]]
ListPlot[{RipleyK[data, Range[0, 200, 10]], 
  RipleyK[proc4, Range[0, 200, 1]]}, 
 PlotLegends -> {"original process", "estimated model"}]
POSTED BY: André Dauphiné
4 Replies
Sort By:

Thanks for your solution. But how I enter "ObservationRegion"

with :

ObservationRegion=Entity["country","France"] or another instruction?

Thanks

POSTED BY: André Dauphiné

You can do:

In[24]:= RipleyK[{proc4, Entity["Country", "France"]}, Range[0, 10, 1]]

Out[24]= {Quantity[0., ("Miles")^2], Quantity[0., ("Miles")^2], 
 Quantity[0., ("Miles")^2], Quantity[0., ("Miles")^2], 
 Quantity[0., ("Miles")^2], Quantity[0., ("Miles")^2], 
 Quantity[0., ("Miles")^2], Quantity[569.827, ("Miles")^2], 
 Quantity[736.026, ("Miles")^2], Quantity[761.407, ("Miles")^2], 
 Quantity[849.261, ("Miles")^2]}
POSTED BY: Gosia Konwerska

The main difficulty is that (AFAIK) there is no exact formula for RipleyK of StraussPointProcess, as well as many others, while it is long known for PoissonPointProcess. The usual workflow is to compare RipleyK of data with RipleyK of (estimated) PoissonPointProcess to investigate if the given point collection exhibits complete spatial randomness. The answer obtained via simulation could be provided though, so we shall look into this.

Update: I looked more carefully into your code and now I know where is the problem. So point statistics will use simulation method for point processes when there is no analytical form. The simulation requires the region input, so the code should be RipleyK[{proc, region}, r], in this case:

RipleyK[{proc4, data["ObservationRegion"]}, Range[0, 200, 1]]
POSTED BY: Gosia Konwerska

Just a guess: Change RipleyK[proc4, Range[0, 200, 1]]}, to RipleyK[proc4, Range[0, 200, 10]]},

POSTED BY: Marvin Ray Burns
Reply to this discussion
Community posts can be styled and formatted using the Markdown syntax.
Tag limit exceeded
Note: Only the first five people you tag will receive an email notification; the other tagged names will appear as links to their profiles.
Reply Preview
Attachments
Remove
or Discard

Group Abstract Group Abstract