R help - Mar 2010 - Frequencies from x/y data into a 2d table (for 3d histogram or heatmap)

Hi all,

I have simple x/y data from screen recording in a sequence:

number,x,y
----------------
1,10,30
1,20,
1,43,110
1,74,18
1,88,112

and would like to create a 3d histogram data structure that i can use
to create a 3d histogram or, more likely a heatmap. The unterlying
data structure therefore needs to look like this:

          0-10        11-20         21-30        31-40
0-10
11-20
21-30
31-40

where the values in the cells represent counts of datapoints fitting
respective ranges for x and y. Can somebody point me to a procedure in
R that can make such a transformation for different range sizes (i.e.
10 in my example)? I came across the following method:

https://stat.ethz.ch/pipermail/r-help/attachments/20070425/fcb92205/attachment.pl


which seems (at least close) but did not get it running (could not
find clear3d()). I also dont' need the 3d histogram plot as I will be
using the table to create a heatmap. I might be thinking to
complicated here. To me, it seems like this should be possible with a
few lines of code. Perhaps one of you has them lying around somewhere
or knows a script that does the trick.

Thanks in advance!

Ralf

Hi Ralf,
have you solved your problem?! If so, could you share? I have the same 
problem...
Best,

Marcio


On 3/25/10 6:03 PM, Ralf B wrote:
> Hi all,
>
> I have simple x/y data from screen recording in a sequence:
>
> number,x,y
> ----------------
> 1,10,30
> 1,20,
> 1,43,110
> 1,74,18
> 1,88,112
>
> and would like to create a 3d histogram data structure that i can use
> to create a 3d histogram or, more likely a heatmap. The unterlying
> data structure therefore needs to look like this:
>
>            0-10        11-20         21-30        31-40
> 0-10
> 11-20
> 21-30
> 31-40
>
> where the values in the cells represent counts of datapoints fitting
> respective ranges for x and y. Can somebody point me to a procedure in
> R that can make such a transformation for different range sizes (i.e.
> 10 in my example)? I came across the following method:
>
>
https://stat.ethz.ch/pipermail/r-help/attachments/20070425/fcb92205/attachment.pl
>
>
> which seems (at least close) but did not get it running (could not
> find clear3d()). I also dont' need the 3d histogram plot as I will be
> using the table to create a heatmap. I might be thinking to
> complicated here. To me, it seems like this should be possible with a
> few lines of code. Perhaps one of you has them lying around somewhere
> or knows a script that does the trick.
>
> Thanks in advance!
>
> Ralf
>