[R] Saved EPS does not match screen when using bquote(.(i))

[Uwe Ligges]

On 23.06.2011 07:14, Dennis Murphy wrote:
> Hi:
>
> As Uwe suggested...
>
> pdf('testgraph.pdf')
> layout( matrix( 1:2 , nrow=2 ) )
> for ( i in 1:2 ) {
>    plot( 0 , 0 , xlab=bquote(mu[.(i)]) )
>   }
> dev.off()
>
> postscript('testgraph.ps')
> layout( matrix( 1:2 , nrow=2 ) )
> for ( i in 1:2 ) {
>    plot( 0 , 0 , xlab=bquote(mu[.(i)]) )
>   }
> dev.off()
>
> png('testgraph.png')
> layout( matrix( 1:2 , nrow=2 ) )
> for ( i in 1:2 ) {
>    plot( 0 , 0 , xlab=bquote(mu[.(i)]) )
>   }
> dev.off()
>
> The three graphs look the same (although the PS graph is rotated to
> landscape while the other two are portrait).

Yes, thanks, and ?postscript will show arguments such as 'horizontal' 
and 'paper' that will help to make them really similar.

Best,
Uwe


 > The main point is that
> mu_1 and mu_2 show up correctly in the two panels in all three graphs
> (at least on my viewers).
>
> The following thread from last January describes some of the problems
> that certain viewers have with Greek letters, which appear to be
> viewer and platform dependent:
>
> http://r-project.markmail.org/search/?q=pdf%20incorrect#query:pdf%20incorrect+page:2+mid:egmb6utulrxgcznw+state:results
>
> I'm guessing that I've seen about a half dozen or so similar posts in
> this forum over the past year and a half, so you can check the list
> archives for related problems.
>
> HTH,
> Dennis
>
> On Wed, Jun 22, 2011 at 8:07 PM, John Kruschke<johnkruschke at gmail.com>  wrote:
>> Here's a fairly minimal-case example in which the saved EPS does not match
>> the screen. The error comes when using bquote(.(i)) instead of bquote(1), as
>> demonstrated by the two minimally different cases below. Very strange. Any
>> clues as to why?
>>
>> #---------------- begin -------------------
>>
>> # Version A. X axis labels have subscripts as constants. EPS is correct.
>> windows()
>> layout( matrix( 1:2 , nrow=2 ) )
>> plot( 0 , 0 , xlab=bquote(mu[1]) )
>> plot( 0 , 0 , xlab=bquote(mu[2]) )
>> savePlot( file="SavePlotTestA.eps" , type="eps" ) # Axis labels are correct
>> in EPS.
>>
>> # Version B. X axis labels have subscripts as variable index. EPS is wrong!
>> windows()
>> layout( matrix( 1:2 , nrow=2 ) )
>> for ( i in 1:2 ) {
>>   plot( 0 , 0 , xlab=bquote(mu[.(i)]) )
>> }
>> savePlot( file="SavePlotTestB.eps" , type="eps" ) # X-AXIS OF PLOT 1 IS
>> WRONG IN EPS.
>>
>> #-------------- end ---------------------
>>
>> Thanks!
>>
>> John K. Kruschke, Professor
>> <http://www.indiana.edu/%7Ekruschke/DoingBayesianDataAnalysis/>
>>
>>
>> 2011/6/22 Uwe Ligges<ligges at statistik.tu-dortmund.de>
>>
>>>
>>>
>>> On 22.06.2011 13:50, John Kruschke wrote:
>>>
>>>> The error happens when using the savePlot() command, like this:
>>>> savePlot( file="TestSavePlot.eps" , type="eps" )
>>>> savePlot( file="TestSavePlot.jpg" , type="jpg" )
>>>>
>>>
>>> Well, plot directly into a device, for postscript:
>>>
>>> postscript("estSavePlot.eps", additionalArguments .....)
>>> plot(1:10)
>>> dev.off()
>>>
>>> Uwe Ligges
>>>
>>>
>>>   The images in the two saved files are not the same, with the JPG being
>>>> correct but the EPS being wrong.
>>>>
>>>> When you suggest "starting separate devices explicitly", what do you mean?
>>>> (I've skimmed through the results of ??device, but can't make sense of
>>>> it.)
>>>> Thank you!
>>>>
>>>> John K. Kruschke, Professor
>>>>
>>>>
>>>> 2011/6/22 Uwe Ligges<ligges at statistik.tu-**dortmund.de<ligges at statistik.tu-dortmund.de>
>>>>>
>>>>
>>>>   I guess you use the menu to save the plots from your Windows device into
>>>>> files rather than starting separate devices explicitly?
>>>>> If so, please use explicit calls to the devices and everything happens as
>>>>> you define it.
>>>>>
>>>>> Uwe Ligges
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> On 22.06.2011 04:31, John Kruschke wrote:
>>>>>
>>>>>   When I save a particular plot as JPG or BMP, the saved image is an
>>>>>> accurate
>>>>>> copy of that plot.
>>>>>> But when I save the same plot as EPS or PDF, some text elements are
>>>>>> incorrectly altered.
>>>>>> An example is attached. Notice in the top middle and top right panels,
>>>>>> the
>>>>>> x-axis labels have correct subscript 1 in the JPG, but incorrect
>>>>>> subscript
>>>>>> 2
>>>>>> in the EPS.
>>>>>>
>>>>>> I'm using R 2.13.0 on Windows 7.
>>>>>>
>>>>>> Any clues regarding the source of this error and its solution would be
>>>>>> appreciated. For example, are there EPS/PDF device drivers that need to
>>>>>> be
>>>>>> separately updated?
>>>>>>
>>>>>> Many thanks.
>>>>>>
>>>>>> John K. Kruschke, Professor
>>>>>> <http://www.indiana.edu/%****7Ekruschke/****DoingBayesianDataAnalysis/<
>>>>>> htt**p://www.indiana.edu/%**7Ekruschke/**DoingBayesianDataAnalysis/<http://www.indiana.edu/%7Ekruschke/DoingBayesianDataAnalysis/>
>>>>>>>
>>>>>>
>>>>>>
>>>>>>>
>>>>>>
>>>>>>
>>>>>> ______________________________****________________
>>>>>> R-help at r-project.org mailing list
>>>>>> https://stat.ethz.ch/mailman/****listinfo/r-help<https://stat.ethz.ch/mailman/**listinfo/r-help>
>>>>>> <https://stat.**ethz.ch/mailman/listinfo/r-**help<https://stat.ethz.ch/mailman/listinfo/r-help>
>>>>>>>
>>>>>>
>>>>>> PLEASE do read the posting guide http://www.R-project.org/**
>>>>>> posting-guide.html<http://www.**R-project.org/posting-guide.**html<http://www.R-project.org/posting-guide.html>
>>>>>>>
>>>>>>
>>>>>> and provide commented, minimal, self-contained, reproducible code.
>>>>>>
>>>>>>
>>>>>
>>>>
>>
>>         [[alternative HTML version deleted]]
>>
>> ______________________________________________
>> R-help at 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 at 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.