Re: [R] plot.mob() fails with cut() error "'breaks' are not unique"

[Achim Zeileis]

On Tue, 22 Jan 2013, Jason Musil wrote:


DeaR all,

I am using mob() for model based partitioning, with a dichotomous 
variable (participant's correct/incorrect response to a test item) 
regressed onto a continuous predictor related to a given property of the 
test item. Although this variable is continuous, the value of this 
variable for many items in this particular analysis is 0. The 
partitioning criterion is self-reported ability in a related area.



mob1 <- mob(

   correct ~ circular.mean | srp.dimension,
   control=mob_control(alpha=.001),
   model=glinearModel,
   family=binomial()
 )


plot(mob1)


Error in cut.default(x, breaks = breaks, include.lowest = TRUE) :
 'breaks' are not unique

The same persists if I specify either a desired number of breaks, or 
explicit breakpoints (e.g. breaks=3 or breaks=c(-0.1,0.1,0.5)). I guess 
this is to do with the funny distribution of the predictor variable, but 
I'm not sure what to do about it.


Jason, you can't pass the "breaks" argument to the plot method directly 
but need to pass it on to the panel function drawing the terminal panels. 
As an example consider


example("mob")
plot(fmPID)
plot(fmPID, tp_args = list(breaks = c(0, 100, 120, 200)))

Hope that fixes your problem.

Best,
Z


Many thanks and apologies if this doesn't fit the mailing list---it is my first 
posting!
Jason Musil

__
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.



__
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

[R] plot.mob() fails with cut() error "'breaks' are not unique"

[Jason Musil]

DeaR all,

I am using mob() for model based partitioning, with a dichotomous variable 
(participant's correct/incorrect response to a test item) regressed onto a 
continuous predictor related to a given property of the test item. Although 
this variable is continuous, the value of this variable for many items in this 
particular analysis is 0. The partitioning criterion is self-reported ability 
in a related area.

> mob1 <- mob(
correct ~ circular.mean | srp.dimension,
control=mob_control(alpha=.001),
model=glinearModel,
family=binomial()
  )

> plot(mob1)

Error in cut.default(x, breaks = breaks, include.lowest = TRUE) : 
  'breaks' are not unique

The same persists if I specify either a desired number of breaks, or explicit 
breakpoints (e.g. breaks=3 or breaks=c(-0.1,0.1,0.5)). I guess this is to do 
with the funny distribution of the predictor variable, but I'm not sure what to 
do about it.

Many thanks and apologies if this doesn't fit the mailing list---it is my first 
posting!
Jason Musil

__
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.