SAS-like Arrays in Stata using Geodist?

[Robert Charles]

I'm trying to find the distance between two lat/long points in Stata. I have 15 lat/long pairs of coordinates in a wide format, and I want to find the distance between (lat1, long1), (lat2, long2) and (lat2, long2), (lat3, long3), etc. Below is an example of two points.

lat1 lon1 lat2 lon2
39.9522613 -75.1949641 39.9482853 -75.1962355

I found the "geodist" function to be useful, so I'd like to use it to loop through everything. I tried the following, but it did not work. Any suggestions?

forval i = 1/15 {
     geodist lat`i' lon`i' lat(`i'+1) lon(`i'+1), gen(km_`i')
}
Was expecting a number or a numeric scalar instead of -lat(1+1)-
r(198);

I also previously had my data in LONG format, with each lat/long coordinate on a separate row, and I tried the following with no luck.

geodist latitude longitude latitude[_n-1] longitude[_n-1], gen(km)

[Leonardo Guizzetti]

Stata doesn't have the same construct as SAS -array-, but it does have local macros which are both more flexible and more (broadly) useful. However, it may take a little more setup to get it working in a similar manner, depending on what you are trying to do. For example, local macros cannot be indexed like a tuple, but there are macro functions that can pull out specific words by position, thus offering a similar mechanism.

To your question, the loop is the way that I would go. The long format approach won't work in this case because -geodist- evidently requires variable names, and that implies a wide-layout.

You were nearly there with your first attempt, but you'll need one more level of nested macros to evaluate the intended arithmetic (i + 1).

forval i = 1/15 {
     geodist lat`i' lon`i' lat`=`i'+1' lon`=`i'+1', gen(km_`i')
}

The macro is resolved from the innermost level, outwards. For example, on the first iteration when i=1:
`=`i'+1' becomes/resolves to ==> `=1+1'
`=1+1' becomes 2

[Robert Charles]

Thank you, Leonardo! That worked!

[Leonardo Guizzetti]

You’re welcome. For a more detailed elaboration, see this Stata FAQ.