Skip to content
Prev 5382 / 29559 Next

Convergence problems with spgwr

Thank you for your replies but it's my understanding that (for an adaptive model) the percentage 
should be entered at the "adapt" argument (based on the spgwr package manual and Roger's reply 
to this posting: https://stat.ethz.ch/pipermail/r-sig-geo/2005-August/000535.html). 
  
Also, the gwr function won't run on my data if I try bandwith = percentage.
This is the error I get:
"Error in print.gwr(list(SDF = <S4 object of class "SpatialPointsDataFrame">,  : 
  subscript out of bounds"

Could you confirm me on this?

-- Adalberto.


Adalberto Pineda
Data Analytics Specialist
MDA Dataquick
9620 Towne Centre Dr.
San Diego, CA. 92121.



-----Original Message-----
From: Roger Bivand [mailto:Roger.Bivand at nhh.no]
Sent: Saturday, March 28, 2009 8:49 AM
To: Corey Sparks
Cc: Adalberto Pineda; r-sig-geo at stat.math.ethz.ch
Subject: Re: [R-sig-Geo] Converge problems with spgwr
On Sat, 28 Mar 2009, Corey Sparks wrote:

            
In addition, the model does have a major scaling problem, as can be seen 
by the CVS scores. Consider re-scaling the variables in the model, and you 
may find that gwr.sel() has an easier time.

Roger