processing dyads in stata

[caleb gwaindepi]

hello Statalisters

I am working with household survey dataset where each row is a subject/observation. In one of the data columns I have an entry with the spouse ID. I want to match couples and characteristics (highest education, gross income, occupational code for primary occupation, perceived health status, to name a few). I created a temporary file where the spousal pid variable is renamed to pid, and the substantive variables are also renamed, and to merge it back into the original data. however i am getting an error message variable pid does not uniquely identify observations in the master data. What can be the problem? If there are polygamous couples in the data how do i detect them?

* Example generated by -dataex-. For more info, type help dataex
clear
input long pid byte(w1_r_edu w1_r_marstt) long w1_r_parhpid byte w1_a_em1occ_isco_c long w1_a_em1inc byte w1_a_hldes
301015 10 1 301043 2 9000 2
301020 6 1 301036 . . 4
301025 12 1 301029 4 -8 1
301026 9 1 301045 . . 4
301027 8 2 501102 . . 1
301029 12 1 301025 5 1500 1
301034 12 1 301035 . . 3
301035 8 1 301034 . . 2
301036 6 1 301020 . . 4
301040 8 1 301017 . . 3
301043 19 1 301015 5 11100 5
301045 10 1 301026 . . 3
301048 12 1 301056 3 -3 3
301052 7 1 301065 . . 4
301056 12 1 301048 2 5800 2
301057 3 1 301058 . . 5
301058 0 1 301057 . . 3
301062 10 1 . . . 1
301064 10 1 501104 . . 1
301065 7 1 301052 . . 4
301067 7 1 310050 . . 1
301073 10 1 301077 8 2000 1
301077 7 1 301073 9 855 3
301079 6 1 301255 . . 3
301080 25 1 301081 . . 1
301081 25 1 301080 . . 4
301084 12 2 301086 4 3150 1
301086 9 2 301084 . . 3
301088 8 1 301092 7 1500 2
301090 10 1 301091 . . 1
301091 6 1 301090 -8 -8 1
301092 12 1 301088 . . 4
301095 10 1 301703 . . 3
301100 11 1 301701 8 2700 1
301102 11 1 501197 . . 1
301104 9 2 302664 . . 4
301105 12 1 301106 2 13768 3
301106 10 1 301105 5 13000 3
301112 22 1 302668 2 -9 1
301117 12 1 309121 2 -8 3
301123 12 2 501467 . . 2
301124 11 1 301125 6 4000 4
301125 7 1 301124 . . 4
301128 17 1 301181 2 7000 3
301134 12 2 301865 7 1995 1
301139 12 1 301151 . . 3
301142 10 1 301147 . . 1
301147 12 1 301142 5 7900 1
301151 19 1 301139 2 9779 4
301154 10 1 301192 . . 1
301156 11 1 301159 . . 5
301157 6 1 301167 . . 5
301164 19 2 301158 2 8000 1
301167 4 1 301157 5 2000 4
301172 20 2 301173 2 9500 1
301173 20 2 301172 2 9500 4
301175 10 1 301176 . . 5
301176 6 1 301175 . . 5
301181 22 1 301128 2 11797 3
301182 12 2 501113 2 2280 2
301196 11 2 301859 7 4000 3
301197 12 1 301198 2 -8 3
301202 11 2 301941 8 5000 1
301204 10 1 301208 5 -3 2
301208 12 1 301204 . . 2
301210 5 2 307343 . . 5
301215 12 1 501292 2 2000 4
301218 22 1 730044 2 28000 2
301219 5 1 . 9 500 1
301220 11 2 301222 . . 2
301222 11 2 301220 . . 2
301223 6 1 301226 . . 5
301224 7 2 301243 7 520 2
301228 11 1 301230 . . 2
301229 25 1 . 7 3000 2
301230 8 1 301228 7 1300 2
301231 8 1 . 7 1200 5
301232 11 1 . 7 3800 2
301235 10 2 301248 8 2000 2
301238 12 1 301239 . . 2
301239 16 1 301238 5 2500 2
301243 10 2 301224 . . 5
301244 9 2 301245 . . 5
301245 10 2 301244 5 1100 3
301247 25 2 301272 . . 5
301248 10 2 301235 . . 5
301253 25 1 501555 5 -9 3
301255 11 1 301079 . . 4
301257 8 2 305515 7 1300 1
301259 11 1 . 8 2700 2
301264 5 1 . . . 2
301269 10 1 . 7 1100 2
301272 6 2 301247 . . 2
301274 12 1 . 9 -9 1
301280 5 1 301961 . . 5
301282 7 1 301294 . . 3
301285 10 2 301947 . . 2
301286 11 2 301287 . . 2
301287 12 2 301286 8 5000 4
301288 25 1 570442 . . 4
end
label values w1_r_edu w1_edgrd
label def w1_edgrd 0 "Grade R/0", modify
label def w1_edgrd 3 "Grade 3/Std. 1", modify
label def w1_edgrd 4 "Grade 4/Std. 2", modify
label def w1_edgrd 5 "Grade 5/Std. 3", modify
label def w1_edgrd 6 "Grade 6/Std. 4", modify
label def w1_edgrd 7 "Grade 7/Std. 5", modify
label def w1_edgrd 8 "Grade 8/Std. 6/Form 1", modify
label def w1_edgrd 9 "Grade 9/Std. 7/Form 2", modify
label def w1_edgrd 10 "Grade 10/Std. 8/Form 3", modify
label def w1_edgrd 11 "Grade 11/Std. 9/Form 4", modify
label def w1_edgrd 12 "Grade 12/Std. 10/Form 5/Matric/Senior Certificate", modify
label def w1_edgrd 16 "Certificate with less than Grade 12/Std. 10", modify
label def w1_edgrd 17 "Diploma with less than Grade 12/Std. 10", modify
label def w1_edgrd 19 "Diploma with Grade 12/Std. 10", modify
label def w1_edgrd 20 "Bachelors degree", modify
label def w1_edgrd 22 "Honours degree", modify
label def w1_edgrd 25 "No Schooling", modify
label values w1_r_marstt w1_marstat
label def w1_marstat 1 "Married", modify
label def w1_marstat 2 "Living with partner", modify
label values w1_r_parhpid w1_parhpid
label values w1_a_em1occ_isco_c w1_occ_code
label def w1_occ_code -8 "Refused", modify
label def w1_occ_code 2 "Professionals", modify
label def w1_occ_code 3 "Technicians and associate professionals", modify
label def w1_occ_code 4 "Clerical support workers", modify
label def w1_occ_code 5 "Service and sales workers", modify
label def w1_occ_code 6 "Skilled agricultural, forestry and fishery workers", modify
label def w1_occ_code 7 "Craft and related trades workers", modify
label def w1_occ_code 8 "Plant and machine operators, and assemblers", modify
label def w1_occ_code 9 "Elementary occupations", modify
label values w1_a_em1inc w1_nonres
label def w1_nonres -9 "Don't know", modify
label def w1_nonres -8 "Refused", modify
label def w1_nonres -3 "Missing", modify
label values w1_a_hldes w1_hldes
label def w1_hldes 1 "Excellent", modify
label def w1_hldes 2 "Very Good", modify
label def w1_hldes 3 "Good", modify
label def w1_hldes 4 "Fair", modify
label def w1_hldes 5 "Poor", modify

code
tempfile original

. save original
file original.dta saved

. drop pid

. rename w1_r_parhpid pid

. rename w1* sp_w1*

. merge 1:1 pid using original
variable pid does not uniquely identify observations in the master data
r(459);

[Clyde Schechter]

merge m:1 pid using original

As an aside, you declare a tempfile named original, but then you never use it. You probably think you used it when you -save original- and later referring to it in the -merge- command. But you did not: the -save original- command created a permanent file named original.dta and the -merge- command made use of it. But it is still cluttering up your hard drive somewhere. To use the tempfile, you need to enclose original in local macro quotes. So the code would look like:

tempfile original // NO MACRO QUOTES HERE
save `original' // MACRO QUOTES NEEDED
...
merge m:1 pid using `original' // MACRO QUOTES NEEDED

[caleb gwaindepi]

Thank you Clyde, it worked