From: Peter Ehlers <ehlers_at_ucalgary.ca>

Date: Wed, 09 Jun 2010 16:30:29 -0600

polygon(c(x1,x2,x3,x4), c(y1,y2,y3,y4), border="blue", lwd=2) polygon(x.rotated, y.rotated, border="red", lwd=2)

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. Received on Wed 09 Jun 2010 - 22:32:36 GMT

Date: Wed, 09 Jun 2010 16:30:29 -0600

On 2010-06-09 14:17, guox_at_ucalgary.ca wrote:

> Thank.

*> Better. Seems that angles are close to but not equal to pi/2.
**> It may be because the plot box is not a square: the length of
**> x-axis is not the same as the length of y-axis.
**> Even curves y = x and y = 1-x look like not orthogonal but
**> they should since multiplication of their slopes is -1.
**> -james
*

I disagree. Your eyes may be fooling you. Setting asp=1 ensures that intervals (x1,x2)=(a,b) and (y1,y2)=(a,b) are of equal length on the plot. (This seems to hold even if you resize the plot after drawing.)

Try removing the box around the plot:

plot(0:30, 0:30, type='n', asp=1, axes=FALSE, xlab="", ylab="")

abline(v=c(0,30), h=c(0,30)) abline(a=0, b=1, col=2) abline(a=30, b=-1, col=4)

polygon(c(x1,x2,x3,x4), c(y1,y2,y3,y4), border="blue", lwd=2) polygon(x.rotated, y.rotated, border="red", lwd=2)

Works for me.

-Peter Ehlers

> plot(1:30,1:30,xlim=c(1,30),ylim=c(1,30),type="n", main = "Rotated

*> rectangle looks like a parallelogram", asp=1)
**> ## a rect at (10,20) with w = 3 and h = 2
**> x = 10
**> y = 20
**> w = 3
**> h = 2
**> x1=x-w
**> y1=y-h
**> x2=x+w
**> y2=y-h
**> x3=x+w
**> y3=y+h
**> x4=x-w
**> y4=y+h
**> polygon(c(x1,x2,x3,x4),c(y1,y2,y3,y4),border="blue")
**>
**> ##Rotate it at (10,20) by 45 degree
**> theta = 45/180*pi
**> x.rotated = c(10 + (x1-10)*cos(theta)-(y1-20)*sin(theta),
**> 10 + (x2-10)*cos(theta)-(y2-20)*sin(theta),
**> 10 + (x3-10)*cos(theta)-(y3-20)*sin(theta),
**> 10 + (x4-10)*cos(theta)-(y4-20)*sin(theta))
**> y.rotated = c(20 + (x1-10)*sin(theta)+(y1-20)*cos(theta),
**> 20 + (x2-10)*sin(theta)+(y2-20)*cos(theta),
**> 20 + (x3-10)*sin(theta)+(y3-20)*cos(theta),
**> 20 + (x4-10)*sin(theta)+(y4-20)*cos(theta))
**>
**> polygon(x.rotated,y.rotated,border="red")
**>
**>> Your transformation assumes that the x- and y-axes are on the
**>> same scale. Add 'asp = 1' to your plot() call to set the
**>> appropriate aspect ratio.
**>>
**>> -Peter Ehlers
**>>
**>>
**>> On 2010-06-09 10:13, guox_at_ucalgary.ca wrote:
**>>> Rectangle R centered at (x,y) with width 2w and height 2h is given by
**>>>
**>>> x1=x-w
**>>> y1=y-h
**>>> x2=x+w
**>>> y2=y-h
**>>> x3=x+w
**>>> y3=y+h
**>>> x4=x-w
**>>> y4=y+h
**>>> polygon(c(x1,x2,x3,x4),c(y1,y2,y3,y4))
**>>>
**>>> Rotating a point (u,v) at (0,0) by theta degree is given by matrix
**>>> [cos(theta),-sin(theta)
**>>> sin(theta),cos(theta)]
**>>> so we have a new point
**>>> (u*cos(theta)-v*sin(theta),u*sin(theta)+v*cos(theta)).
**>>>
**>>> Hence rotated R by theta at (x,y) is given by
**>>>
**>>> x.rotated = c(x + (x1-x)*cos(theta)-(y1-y)*sin(theta),
**>>> x + (x2-x)*cos(theta)-(y2-y)*sin(theta),
**>>> x + (x3-x)*cos(theta)-(y3-y)*sin(theta),
**>>> x + (x4-x)*cos(theta)-(y4-y)*sin(theta))
**>>> y.rotated = c(y + (x1-x)*sin(theta)+(y1-y)*cos(theta),
**>>> y + (x2-x)*sin(theta)+(y2-y)*cos(theta),
**>>> y + (x3-x)*sin(theta)+(y3-y)*cos(theta),
**>>> y + (x4-x)*sin(theta)+(y4-y)*cos(theta))
**>>>
**>>> polygon(x.rotated,y.rotated)
**>>>
**>>> But it turns out to be a parallelogram with angles not equal to 90,
**>>> not a rectangle. See R code below.
**>>>
**>>> Any way to improve this so that the rotated rectangle looks like a
**>>> rectangle? Thanks,
**>>>
**>>> -james
**>>>
**>>>
**>>> plot(1:10,1:10,xlim=c(1,20),ylim=c(1,40),type="n", main = "Rotated
**>>> rectangle looks like a ")
**>>> ## a rect at (10,20) with w = 3 and h = 2
**>>> x = 10
**>>> y = 20
**>>> w = 3
**>>> h = 2
**>>> x1=x-w
**>>> y1=y-h
**>>> x2=x+w
**>>> y2=y-h
**>>> x3=x+w
**>>> y3=y+h
**>>> x4=x-w
**>>> y4=y+h
**>>> polygon(c(x1,x2,x3,x4),c(y1,y2,y3,y4),border="blue")
**>>>
**>>> ##Rotate it at (10,10) by 45 degree
**>>> theta = 45/180*pi
**>>> x.rotated = c(10 + (x1-10)*cos(theta)-(y1-20)*sin(theta),
**>>> 10 + (x2-10)*cos(theta)-(y2-20)*sin(theta),
**>>> 10 + (x3-10)*cos(theta)-(y3-20)*sin(theta),
**>>> 10 + (x4-10)*cos(theta)-(y4-20)*sin(theta))
**>>> y.rotated = c(20 + (x1-10)*sin(theta)+(y1-20)*cos(theta),
**>>> 20 + (x2-10)*sin(theta)+(y2-20)*cos(theta),
**>>> 20 + (x3-10)*sin(theta)+(y3-20)*cos(theta),
**>>> 20 + (x4-10)*sin(theta)+(y4-20)*cos(theta))
**>>>
**>>> polygon(x.rotated,y.rotated,border="red")
**>>>
**>>>
**>>>> On 06/04/2010 01:21 AM, guox_at_ucalgary.ca wrote:
**>>>>> boxed.labels draw text with box well.
**>>>>> But, the box cannot be shadowed and srt = 45 seems not to work:
**>>>>> text is rotated but the box does not.
**>>>>> polygon.shadow can rotate and shadow but have to calculate its
**>>>>> dimensions,
**>>>>> based on the text length and size.
**>>>>> Do you have any other way to draw text with rotated and shadowed box?
**>>>>
**>>>> The srt argument was intended to allow the user to rotate the text in
**>>>> 90
**>>>> degree increments, and the box just changes shape to fit whatever is
**>>>> in
**>>>> it. The underlying function that draws the box (rect) doesn't have a
**>>>> rotation argument. It would be possible to write a special function
**>>>> using polygon, just do the calculations for box size and then rotate
**>>>> the
**>>>> text with srt= and the polygon by transforming the coordinates of the
**>>>> vertices, as long as the default justification (center) is used. I
**>>>> can't
**>>>> do this right at the moment, but if you are really stuck I might be
**>>>> able
**>>>> to do it in the near future.
**>>>>
**>>>> Jim
**>>>>
**>>>>
*

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. Received on Wed 09 Jun 2010 - 22:32:36 GMT

Archive maintained by Robert King, hosted by
the discipline of
statistics at the
University of Newcastle,
Australia.

Archive generated by hypermail 2.2.0, at Wed 09 Jun 2010 - 23:20:28 GMT.

*
Mailing list information is available at https://stat.ethz.ch/mailman/listinfo/r-help.
Please read the posting
guide before posting to the list.
*