RE: st: Count multiple observations / weight in Scatter

[Nick Cox <n.j.cox@durham.ac.uk>]

For just counting, -contract- is easier and incidentally faster than -collapse-. 

An alternative to Friedrich's code is 

drop _all
input id x y
1 10 19
2 12 19
3 10 19
end
contract x y
list
scatter y x [aw=_freq]

Nick 
n.j.cox@durham.ac.uk 

Friedrich Huebler
=================

Maarten's code generates the data Rainer had asked for but I would
suggest to -collapse- the data instead. A scatter plot of the original
data is inefficient because observations 1 and 3 will be plotted
twice, on top of each other. The information in the variable "id" is
lost at that point so it is not necessary to keep it for the graph.

drop _all
input id x y
1 10 19
2 12 19
3 10 19
end
gen count = 1
collapse (sum) count, by(x y)
list
scatter y x [aw=count]

Maarten buis
============

> *--------------- begin example -------------
> drop _all
> input id       x            y
> 1             10           19
> 2             12           19
> 3             10           19
> end
> bys x y : gen count = _N
> sort id
> list
> *--------------- end example ----------------
> (For more on examples I sent to the Statalist see:
> http://www.maartenbuis.nl/example_faq )

Rilke Rainer Michael
====================

>> I have data where for example one subject indicates two
>> different variables, lets say X and Y.
>> I now want to create a variable lets say freq which counts
>> the number of multiple duplicate observation.


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/