R help - Apr 2009 - plotCI (plotrix) problem

I am attempting to create a plot with intervals "stretched" in the
x-direction using plotCI() in the plotrix package.  The same data
provides an appropriate set of intervals when "stretched" in the
y-direction but I only get a lower interval when "stretched" in the
x-direction.

 

The data are as follows

 

mns <-
c(-0.00584,-0.00458,-0.00934,-0.00969,-0.01007,0.01146,-0.00261,-0.00287
,0.00035,-0.00248,0.00084,0.01183,-0.00617,-0.00789,-0.00405,-0.01076,0.
00285,0.01464,0.01665,0.00179,-0.00038,-0.00046,0.00388,0.00501,0.00237,
-0.00488,-0.00081,-0.00432,-0.00172,0.00574,0.00853,0.00869,-0.00551,-0.
00069,0.00047)

ses <-
c(0.001552,0.001341,0.001442,0.001511,0.001794,0.001735,0.001288,0.00136
4,0.001826,0.00196,0.001037,0.00009,0.00153,0.00147,0.002195,0.001189,0.
002442,0.001293,0.000803,0.001359,0.001936,0.001431,0.001962,0.002132,0.
001879,0.002846,0.003047,0.003176,0.003178,0.00328,0.001725,0.003523,0.0
03681,0.002316,0.002367)

 

This plot works as expected ...

 

require(plotrix)

plotCI(1:length(mns),mns,ses)

 

... but this one does not ...

 

plotCI(mns,1:length(mns),ses,err="x")

 

If I multiply the ses values by successively larger values then more of
the points are plotted with the upper portion of the interval, e.g., ...

 

plotCI(mns,1:length(mns),10*ses,err="x")

 

which implies that this is a scaling issue.  >From this, I suspect (am
guessing) that the issue arises somewhere in this part of the plotCI()
code ...

 

        nz <- abs(li - pmax(x - gap, li)) * x.to.in > 0.001

        ...

        nz <- abs(ui - pmin(x + gap, ui)) * y.to.in > 0.001

 

but I don't quite understand what is going on here (and elsewhere in
plotCI()) and my "guesses" at corrections did not work.  My searches
on
R-help were not fruitful.

 

So, does anyone have any suggestions?  Thanks in advance for any help.

 

 

Note my session information below ...

 
> sessionInfo()
R version 2.8.0 (2008-10-20) 

i386-pc-mingw32 

 

locale:

LC_COLLATE=English_United States.1252;LC_CTYPE=English_United
States.1252;LC_MONETARY=English_United
States.1252;LC_NUMERIC=C;LC_TIME=English_United States.1252

 

attached base packages:

[1] stats     graphics  grDevices datasets  tcltk     utils     methods
base     

 

other attached packages:

[1] plotrix_2.5-4  svSocket_0.9-5 TinnR_1.0.2    R2HTML_1.59
Hmisc_3.5-2   

 

loaded via a namespace (and not attached):

[1] cluster_1.11.12 grid_2.8.0      lattice_0.17-20 svMisc_0.9-5
tools_2.8.0

 

 

 


	[[alternative HTML version deleted]]

Derek Ogle <DOgle <at> northland.edu> writes:
> 
> I am attempting to create a plot with intervals "stretched" in
the
> x-direction using plotCI() in the plotrix package.  The same data
> provides an appropriate set of intervals when "stretched" in the
> y-direction but I only get a lower interval when "stretched" in
the
> x-direction.
> 
> The data are as follows
..

This looks  like a bug to me, and you should contact the author.
As a temporary workaround, use plotCI from package gplots which is
one father (the grandfather is Bill Venables).

Dieter

Derek Ogle <DOgle <at> northland.edu> writes:
> 
> I am attempting to create a plot with intervals "stretched" in
the
> x-direction using plotCI() in the plotrix package.  The same data
> provides an appropriate set of intervals when "stretched" in the
> y-direction but I only get a lower interval when "stretched" in
the
> x-direction.
> 
>         nz <- abs(ui - pmin(x + gap, ui)) * y.to.in > 0.001
It's a copy-paste typo. The above line should be 
        nz <- (abs(ui - pmin(x + gap, ui)) * x.to.in) > 0.001

Dieter