I am trying to find the nearest neighbour using the geonear command conditional on observations sharing the same group. For example, matching individuals to their nearest neighbour (in kms) who shares the same gender. Below is an example of trying to match "cell towers" that share the same group identifier. This is adapted from the description in the geonear help file to match cell towers based on their assigned group. Is there a more efficient way to do this other than what I am doing below?
** Create Dataset
clear
set obs 1000
gen ctid = _n
gen double ctlat = 37 + (41 - 37) * uniform()
gen double ctlon = -109 + (109 - 102) * uniform()
sort ctid
g ctgroup = ceil(ctid / 10) // group cell towers into groups of 10
sum ctgroup
local G = `r(max)'
tempfile cell
save `cell',replace
** Use geonear withn each group
forvalues j=1(1)`G'{
use `cell' , clear
keep if ctgroup == `j'
tempfile cell`j'
save `cell`j'' , replace
geonear ctid ctlat ctlon using `cell`j'', n(ctid ctlat ctlon) ignoreself
save `cell`j'' , replace
}
** Append the per group datasets
use `cell1'
forvalues j = 2(1)`G'{
append using `cell`j''
}
]
** Create Dataset
clear
set obs 1000
gen ctid = _n
gen double ctlat = 37 + (41 - 37) * uniform()
gen double ctlon = -109 + (109 - 102) * uniform()
sort ctid
g ctgroup = ceil(ctid / 10) // group cell towers into groups of 10
sum ctgroup
local G = `r(max)'
tempfile cell
save `cell',replace
** Use geonear withn each group
forvalues j=1(1)`G'{
use `cell' , clear
keep if ctgroup == `j'
tempfile cell`j'
save `cell`j'' , replace
geonear ctid ctlat ctlon using `cell`j'', n(ctid ctlat ctlon) ignoreself
save `cell`j'' , replace
}
** Append the per group datasets
use `cell1'
forvalues j = 2(1)`G'{
append using `cell`j''
}
]
Comment