# st: Re: Nearest distance (spatial) and shp2dta question

 From "David Torres" To Subject st: Re: Nearest distance (spatial) and shp2dta question Date Fri, 3 Jul 2009 09:43:24 -0400

```Sergio,

Thank you for this.  I will certainly try this and let you know if it worked
for what I want.

Cheers,
David Diego Torres

----- Original Message -----
To: <statalist@hsphsun2.harvard.edu>
Sent: Friday, July 03, 2009 9:40 AM
Subject: st: Nearest distance (spatial) and shp2dta question

> Here another resource (which I got before from the kind people at Stata):
>
> Havershine is a user written command that calculates the distance using
> the Haversine formula, Vincenty formula, and/or the Law-of-Cosines-based
> formula. Type the following to install this command:
>
> net describe vincenty, from(http://fmwww.bc.edu/RePEc/bocode/v)
> net install vincenty
>
> Then, type -help vincenty- to view the help file. By using the hav()
> option you
> can calculate the distance using the Haversine formula instead of manually
> entering the equation. Also, the default is to express the distances in
> miles
> but you can change this to kilometers with the in km option.
>
> See one example here
>
> *********************
> use testdistance, clear
> set more off
> sort patient
> gen id = _n
> qui levelsof patient, local(patients)
> qui gen shortest_dist = .
> qui gen matched_hos = .
> foreach p of local patients {
>         preserve
>         qui keep if patient==`p'
>         local pat_long = pat_long[1]
>         local pat_lat = pat_lat[1]
>         local obs = id[1]
>         restore
>         vincenty `pat_lat' `pat_long' hos_lat hos_long, hav(h) replace
>         summarize h
>         qui replace shortest_dist = `r(min)' in `obs'
>         preserve
>         keep if round(h,.00001) == round(`r(min)', .00001)
>         local id = hospital[1]
>         restore
>         qui replace matched_hos = `id' in `obs'
>
> }
> drop id
> ***********************
>
> This is calculating the straight distance from the patient to the
> hospital. This is not taking into consideratin any physical land barriers,
> such as mountains and rivers.
>
>
> --
> Sergio I. Prada, PhD (Candidate)
> Department of Public Policy
> University of Maryland, Baltimore County
> 1000 Hilltop Circle, Baltimore, MD 21250
> Phone: 410-455-6527
>
> *
> *   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/
>
*
*   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/
```