Re: Legends

From: Rosemarie Drummond <rmarie_at_nyahnyahspammersnyahnyah>
Date: Mon, 10 Nov 2008 22:15:16 -0400

Hi David,
  Thankyou so much for your help!
I have now got it to do something reasonable using regular legend labels
with lgBoxMinorExtentF=0.25 and lgAutoManage=False.
I had realised the problem with the line properties was connected with
the Boxlines which you rightly guessed I had on only to show the space
actually occupied by the legends.
Thanks!
Rosemarie

David Brown wrote:
>
> To follow up, I did indeed find a problem that involved the line
> buffer not being flushed properly in all cases. This only shows up in
> the Legend when
> the lgBoxLinesOn is set True and the lgBoxLineDashPattern is set to
> "SolidLine". So presuming that you only have the box lines turned on
> in order to
> show the space occupied by the individual Legend items, this will not
> be a problem for you. However, in a few days, we can give you a new
> binary if
> you need it.
>
> Also I forgot to mention another resource you might want to play with:
> lgLineLabelDashSegLenF. This specifies the length of dash segment
> before it starts to repeat.
> Since the line labels are actually part of the dash pattern, if the
> dash segment length is too short, you may see multiple instances of
> the label within the legend item.
> -dave
>
>
> On Nov 10, 2008, at 5:48 PM, David Brown wrote:
>
>> Hi Rosemarie,
>>
>> It would be helpful to know what you ultimately want your legend to
>> look like. Do you really want it to be as wide as your script suggests?
>> Why are you not using the regular legend labels to label your lines?
>> Because you couldn't figure out how to set the size? Or was it
>> something else, like wanting the label color to match the line color?
>>
>> Some resources that you may not know about are:
>>
>> lgAutoManage: when set True, the Legend label font size is set
>> automatically. Setting this False allows you to control the label
>> font sizes.
>> It does not affect the line label font size (these are actually part
>> of the line pattern).
>>
>> lgBoxMinorExtentF : this basically controls the fraction of the
>> legend as a whole that the boxes occupy in the direction orthogonal
>> to the sequence of legend lines.
>> By default it is .5, to leave plenty of room for the legend labels.
>> However, if you have turned the labels off, setting this resource to
>> a larger value will cause
>> the boxes to be wider (assuming the legend lines oriented in a
>> vertical sequence). If you set it to 1.0, then the boxes will occupy
>> the full width of the space allotted by
>> vpWidthF if you also set lgLeftMarginF and lgRightMarginF to 0.0.
>>
>> lgBoxMajorExtentF: this controls the spacing between each box (in the
>> vertical direction, given the orientation you have set up for your
>> legend).
>>
>> I hope this helps. As I stated above, it would be helpful to have a
>> better sense of what you want your ultimate result to look like.
>> -dave
>>
>> P.S. I am noticing an anomaly in your legend that could be a bug in
>> the code. When I run your code, the second two lines appear not to
>> restore all the line properties
>> after the line label text is drawn. I will look into this.
>>
>>
>> On Nov 10, 2008, at 2:24 PM, Rosemarie Drummond wrote:
>>
>>> Hi,
>>> I am trying to add a legend to a plot. Below is the ncl script just
>>> for this legend part. I am attaching a ps file created by it.
>>> I have put lgBoxLinesOn = True so that it shows how much white space
>>> it puts each side of the actual legends.
>>> I would like this white space to be very small and the text to be
>>> bigger. The actual line segments could be shorter but that does not
>>> seem to be settable.
>>> Any advice would be gladly received
>>> Rosemarie
>>>
>>>
>>>
>>>
>>> load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_code.ncl"
>>> load "$NCARG_ROOT/lib/ncarg/nclscripts/csm/gsn_csm.ncl"
>>> begin
>>> ncolors = 9
>>> cmap = new((/ncolors,3/), float)
>>> cmap = (/ (/1.0,1.0,1.0/), \
>>> (/0.0,0.0,0.0/), \
>>> (/1.0,0.0,0.0/), \
>>> (/0.0,1.0,0.0/), \
>>> (/0.0,0.0,1.0/), \
>>> (/0.0,1.0,1.0/), \
>>> (/1.0,0.0,0.80/), \
>>> (/1.0,0.50,0.0/), \
>>> (/0.90,0.90,0.90/)/)
>>>
>>> xworkid = gsn_open_wks("ps","Vtst")
>>>
>>> setvalues xworkid
>>> "wkColorMap" : cmap
>>> end setvalues
>>> getvalues xworkid
>>> "wkDefGraphicStyleId": gsid
>>> end getvalues
>>>
>>>
>>> Lg_id = create "Legend" legendClass xworkid
>>> "vpXF" : 0.3
>>> "vpYF" : 0.5
>>> "vpWidthF" : 0.70
>>> "vpHeightF" : 0.18
>>> "lgItemCount" : 3
>>> ; "lgPerimOn" : False
>>> "lgBoxLinesOn" : True
>>> "lgLeftMarginF" : 0.01
>>> "lgRightMarginF" : 0.01
>>>
>>> "lgLabelsOn" : False
>>> "lgLabelFont" : 30
>>> "lgLabelFontHeightF" : .07
>>> "lgLabelStrings" : (/" D:F35:)):F30: Models"," VM5a"," VM2"/)
>>>
>>> "lgDashIndexes" : (/16,0,0/)
>>> "lgMonoLineThickness" : False
>>> "lgLineThicknesses" : (/3.,6.,6./)
>>> "lgLineColors" : (/4,3,2/)
>>>
>>> ; "lgMonoLabelFontColor" : False invalid
>>> ; "lgLabelFontColors" : (/4,3,2/) invalid
>>>
>>> "lgMonoLineLabelFontHeight" : False
>>> "lgLineLabelFontHeights" : (/0.02,0.02,0.02/)
>>> "lgLineLabelFontThicknessF" : 3
>>> "lgLineLabelFontColors" : (/4,3,2/)
>>> "lgLineLabelStrings" : (/"VM5b-#","VM5a ","VM2 "/)
>>> "lgLineLabelFont" : 30
>>>
>>> end create
>>> draw(Lg_id)
>>>
>>> end
>>>
>>> --
>>> Dr Rosemarie Drummond
>>> rmarie_at_atmosp.physics.utoronto.ca
>>>
>>> <Vtst.ps>_______________________________________________
>>> ncl-talk mailing list
>>> List instructions, subscriber options, unsubscribe:
>>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>
>> _______________________________________________
>> ncl-talk mailing list
>> List instructions, subscriber options, unsubscribe:
>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>

-- 
Dr Rosemarie Drummond
rmarie_at_atmosp.physics.utoronto.ca
_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Mon Nov 10 2008 - 19:15:16 MST

This archive was generated by hypermail 2.2.0 : Fri Nov 14 2008 - 16:42:12 MST