[R] why the length and width of a plot region produced by the dev.new() function cannot be correctly set?

David Winsemius dw|n@em|u@ @end|ng |rom comc@@t@net
Fri May 4 18:26:42 CEST 2018


> On May 4, 2018, at 12:04 AM, sunyeping <sunyeping using aliyun.com> wrote:
> 
> 
> ------------------------------------------------------------------
> From:David Winsemius <dwinsemius using comcast.net>
> Send Time:2018 May 4 (Fri) 13:25
> To:孙业平 <sunyeping using aliyun.com>
> Cc:R Help Mailing List <r-help using r-project.org>
> Subject:Re: [R] why the length and width of a plot region produced by the dev.new() function cannot be correctly set?
> 
> 
> > On May 3, 2018, at 6:28 PM, sunyeping via R-help <r-help using r-project.org> wrote:
> > 
> > When I check the size of the plot region usingdev.size("in")a new plot region is produced and in the Rconsole I get[1] 5.333333 5.322917
> 
> Your test is all mangleed together. You failed in your duty to read the list info and the Posting guide . NO HTML!
> 
> > If I mean to produce a plot region with size setting bydev.new(length=3,width=3)a plot region is produced, but the size is [2.281250, 5.322917], as detected by the de.size function. If I type:dev.new(length=10,width=10)I get a plot region of with the size of [7.614583, 5.322917]. It seems that the width of the new plot region cannot be set, and tt is always 5.322917. The length of the new plot region can be set, but it is always smaller that the values I set.What do I miss? What is the correct way of setting the dimension of the new plot region? I will be grateful to any help.Best regards,
> 
> The size of the device is not the size of the plot region. You need to take into account the margins. See ?par
> 
> Thank you, David.
> I have read the par() document. Clearly the size of the plot region is smaller than or equal to the divice size. However, if I produce a graphic device with dev.new (length, width) or other functions, I find the largest  width of the new device is always 5.3 inches whatever the values I set, and the length of it is alway smaller than what I set. Could you tell me how to produce a graphic divice with correct size that I set? I need this function because the graphic divice cannot accomendate all of the graph I make with some of plot tools such as ggtree. In ggtree plot, part of the tree tips label are invisible (https://www.dropbox.com/s/87gyusx7ay1xxu8/tree.pdf?dl=0) even I set "par(mar=rep(0,4))". So I think I must plot the tree on a larger graphic device.  

- The argument to give dev.new is 'height', not 'length'.

- There are two margins, an inner margin (controlled by `mar`) and an outer margin (controlled by `oma`). 

- The link to the dropbox image does demonstrate that you are having problems with your ggplot efforts, but it wasn't a link to any test data or the code used to produce that tree, so I'm not in a position to do any testing to offer refinements. From the red/bold highlighting in the HTML copy I got directly (that none of the other viewers are seeing)  I can see you are frustrated, but you need to take responsibility for writing a message that is being mangled. We readers of the this plain-text-only mailing list were seeing it. 

- Please read the Posting Guide about the need to information about your specific computer setup and the value of a reproducible example. I suggest you should be learning to use the png or pdf devices. Read ?Devices and ?png. (The default units are not inches.).


> Best regards. 
> 
> 
> > 
> >  [[alternative HTML version deleted]]

- It appears that your second message was also sent as HTML. I do therefore ask (again) that you take the time to read the Posting Guide.

\/\/\/\/A message that appears in every message sent from Rhelp's list server \/\/\/\/\/\/\
> > PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

David Winsemius
Alameda, CA, USA

'Any technology distinguishable from magic is insufficiently advanced.'   -Gehm's Corollary to Clarke's Third Law




More information about the R-help mailing list