At first I thought this is a problem for -cluster-, but it
doesn't have quite the usual structure. Ken Higbee may be
able to tell us how to use that.
You can put in a single observation the ids
of all nests within 100 units of any given nest like this:
I assume variables nest_id1, nest_id2, distance.
gen near_ids = ""
bysort nest_id1 (distance) : replace near_ids = near_ids[_n-1] + string(nest_id2) + " " if distance <= 100
by nest_id1 : replace near_ids = trim(near_ids[_N])
If your ids are already string, you won't need the -string()-.
If you have many more near nests than can fit into a string variable
in your Stata, this is no use. Otherwise it seems the first step
in your problem.
Nick
[email protected]
> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]]On Behalf Of
> Anirudh V. S.
> Ruhil
> Sent: 09 December 2006 23:36
> To: [email protected]
> Subject: st: distance-based grouping
>
>
> Given the following data:
>
> nest_id nest_id distance
> 0 408 67.38784
> 0 35 55.74133
> 0 299 72.978
> 0 496 96.11179
> 0 373 62.25668
> 0 466 32.23399
> 0 358 52.49713
> 0 0 0
> 0 42 90.56657
> 0 407 75.57355
> 1 161 21.8282
> 1 128 30.91449
> 1 373 91.10458
> 1 465 54.24871
> 1 496 68.08784
> 1 1 0
> 1 42 61.35756
> 1 34 21.97413
>
> where "nest_id" flags unique bird nests, and "distance" is
> measured in
> feet.
>
> I need to identify all nests within a specified distance (for
> e.g., 100) of
> each other, and thus create unique clusters of nests. These
> clusters will
> contain nests that show up once per cluster but may show up
> in more than
> one cluster (for e.g., nest 496 shows up in the same cluster
> as nest 0,
> shows up again in a different cluster with nest 1, but nest 0
> and nest 1
> are not in a common cluster). Hopefully my articulation of
> the problem is
> comprehensible, and someone has a clue as to the appropriate
> syntax. I need
> to be able to run this code for nests along several miles of
> a river so
> anything that cuts down on manual labor would be invaluable.
>
> best
>
> Ani
> *
> * For searches and help try:
> * http://www.stata.com/support/faqs/res/findit.html
> * http://www.stata.com/support/statalist/faq
> * http://www.ats.ucla.edu/stat/stata/
>
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
