On Thu, 4 Oct 2001 14:04:53 +0200 (MET DST),
Agustin Lobo (AL) wrote:
> Hi! I'm testing pixmap. I've found the following:
>> dim(lissN543)
> [1] 270 900 3
>> a <- pixmap(lissN543/255,ncol=900,nrow=270,type="rgb")
>> dim(a)
> [1] 270 900 3
>>
>> plot.pixmap(a)
> ...and I get the image as a vertical rectangle (I was expecting an
> horizontal rectangle). Is this an error or am I doing something
> wrong?
You have 900 columns and only 270 rows, hence you should have a
vertical rectangle that is approx. 3 times as wide as high.
The difference to calling image() on a matrix is deliberate: if you
store a picture as a matrix (as usual), then you'd expect element 1,1
of that matrix to correspond to the upper left corner. That is
different from what image() does, because it was designed for doing
the right thing for matrices produced by, e.g., outer() ... [at least
I think that's the reason for image() rotating the matrix].
Best,
Fritz
-------------------------------------------------------------------
Friedrich Leisch
Institut für Statistik Tel: (+43 1) 58801 10715
Technische Universität Wien Fax: (+43 1) 58801 10798
Wiedner Hauptstraße 8-10/1071 Friedrich.Leisch@ci.tuwien.ac.at
A-1040 Wien, Austria http://www.ci.tuwien.ac.at/~leisch
-------------------------------------------------------------------
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !) To: r-devel-request@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
a <- pixmap(lissN543/255,ncol=900,nrow=270,type="rgb")
dim(a)
[1] 270 900 3
plot.pixmap(a)
...and I get the image as a vertical rectangle (I was expecting an
horizontal rectangle). Is this an error or am I doing something
wrong?
Agus
Dr. Agustin Lobo
Instituto de Ciencias de la Tierra (CSIC)
Lluis Sole Sabaris s/n
08028 Barcelona SPAIN
tel 34 93409 5410
fax 34 93411 0012
alobo@ija.csic.es
On Thu, 6 Sep 2001, Friedrich Leisch wrote:
There is a new package pixmap in $CRAN/src/contrib/Devel which
provides a first shot at a class hierarchy for bitmapped images (as
both bitmap and image are already taken I named it pixmap ... if
somebody has an idea for a better name, please let me know).
This is by no means meant to be final ... more like a basis for
discussion about classes for that purpose. It already has RGB, indexed
and grey bitmaps and conversion and plotting methods for those.
Thanks to Roger Bivand there are native R functions for
reading/writing PNM images, this replaces the respective functions in
the e1071 package. Hence, the netpbm libraries are no longer needed,
which should make life easier (especially on windows). Once we have
agreed on the basic classes it should be easy to support import/export
of other image formats like jpeg or png.
All feedback & code is welcome!
Best,
Fritz
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !) To: r-devel-request@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
You have 900 columns and only 270 rows, hence you should have a
vertical rectangle that is approx. 3 times as wide as high.
The difference to calling image() on a matrix is deliberate: if you
store a picture as a matrix (as usual), then you'd expect element 1,1
of that matrix to correspond to the upper left corner. That is
different from what image() does, because it was designed for doing
the right thing for matrices produced by, e.g., outer() ... [at least
I think that's the reason for image() rotating the matrix].
I agree with the point that the upper left corner must be the
1,1 cell in the matrix for a "natural" display of the matrix
as an image. But this does not mean that the matrix should be rotated.
In all image processing packages that I know of, an image with
200 cols and 100 rows is displayed as a horizontal rectangle. Note
that, if there is geometric correction to a geographic projection,
the columns go W to E as left to right and the rows N to S as
top to bottom. In fact,
the "natural" display should follow the same geometry than print,
i.e.,
should be displayed as a horizontal rectangle.
Of course you can do it as you want because this is your package,
but are you sure that you want it the way it is?
Regards,
Agus
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !) To: r-devel-request@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
--RL6cCFzpfl
Content-Type: text/plain; charset=us-ascii
Content-Description: message body text
Content-Transfer-Encoding: 7bit
On Thu, 4 Oct 2001 14:29:58 +0200 (MET DST),
Agustin Lobo (AL) wrote:
> On Thu, 4 Oct 2001, Friedrich Leisch wrote:
[...]
> I agree with the point that the upper left corner must be the
> 1,1 cell in the matrix for a "natural" display of the matrix
> as an image. But this does not mean that the matrix should be rotated.
> In all image processing packages that I know of, an image with
> 200 cols and 100 rows is displayed as a horizontal rectangle. Note
> that, if there is geometric correction to a geographic projection,
> the columns go W to E as left to right and the rows N to S as
> top to bottom. In fact,
> the "natural" display should follow the same geometry than print,
> i.e.,
>> matrix(1:12,ncol=6,nrow=2)
> [,1] [,2] [,3] [,4] [,5] [,6]
> [1,] 1 3 5 7 9 11
> [2,] 2 4 6 8 10 12
>>
> should be displayed as a horizontal rectangle.
Excactly, I agree with everything you say!
you mean it doesn't do that? which version of pixmap, which version of
R on which platform?
Using
R> x <- matrix(1:12,ncol=6,nrow=2)
R> library(pixmap)
R> plot(as.pixmap(x))
I get the attached picture, which is what you want, no?
Best,
Fritz
--RL6cCFzpfl
Content-Type: image/png
Content-Disposition: inline;
filename="pixmap.png"
Content-Transfer-Encoding: base64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--RL6cCFzpfl--
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !) To: r-devel-request@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
ok, found the pronlem: my fault.
I was using v.0.1. With Version: 0.2-1
it's ok.
I observe that you stretch lineraly now within
pixmap. The problem is that you use the max
of all 3 bands, which is not normally good.
Also an optional "normal" stretching should
provide better results in most cases. I'll
send you the code and an attached example
(only to Friedrich.Leisch@ci.tuwien.ac.at to avoid
annoying other people) in a while.
Agus
Dr. Agustin Lobo
Instituto de Ciencias de la Tierra (CSIC)
Lluis Sole Sabaris s/n
08028 Barcelona SPAIN
tel 34 93409 5410
fax 34 93411 0012
alobo@ija.csic.es
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !) To: r-devel-request@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._