Compute Female proportion in each district in long format data in Stata.

[smith Jason]

I have a long format dataset where it has 5 variables:
did---district id
sid---school id
id--student id
age --student age
count-Bonus count
male--student gender(1-male;0-female)
Now, I want to compute female proportions for each district.
Thank you for your Stata code.


* Example generated by -dataex-. For more info, type help dataex.
clear
input int did long sid float(id age count male)
1 11 111 42 20 1
1 11 111 30 20 1
1 11 111 42 10 1
1 11 111 34 10 1
1 11 111 30 10 1
1 11 111 33 10 1
1 12 112 39 20 0
1 12 112 40 20 0
1 12 112 43 20 0
1 12 112 33 20 0
1 13 113 40 10 1
1 13 113 41 20 1
1 13 113 38 20 1
1 14 114 32 10 0
1 14 114 35 10 0
1 14 114 32 20 0
1 14 114 37 10 0
1 15 115 31 10 0
1 15 115 29 10 0
1 15 115 36 10 0
1 15 115 37 20 0
1 16 116 31 20 1
1 16 116 36 20 1
1 16 116 38 10 1
1 16 116 27 10 1
1 16 116 35 20 1
1 17 117 39 10 0
1 17 117 44 20 0
1 17 117 41 10 0
1 17 117 34 20 0
1 17 117 45 20 0
1 17 117 28 10 0
2 21 211 48 10 0
2 21 211 38 10 0
2 21 211 31 20 0
2 21 211 34 10 0
2 22 222 42 10 1
2 22 222 37 20 1
2 22 222 40 20 1
2 22 222 44 10 1
2 22 222 42 20 1
2 23 222 37 10 1
2 23 223 43 20 1
2 23 223 47 10 1
2 23 223 35 20 1
2 24 224 44 20 0
2 24 224 40 10 0
2 24 224 45 10 0
2 24 224 36 20 0
2 24 224 39 10 0
2 25 225 41 20 1
2 25 225 34 20 1
2 25 225 39 20 1
2 25 225 36 10 1
2 25 225 38 20 1
2 26 226 33 10 0
2 26 226 46 10 0
2 26 226 45 20 0
2 26 226 35 10 0
2 26 226 41 10 0
2 27 227 33 20 1
2 27 227 43 10 1
2 27 227 46 20 1
2 27 227 32 20 1
3 31 331 35 20 0
3 31 331 42 10 0
3 32 332 73 10 1
3 32 332 29 20 1
end

[Ken Chui]

Now, I want to compute female proportions for each district.

^ This is a little vague and I am not sure what exactly is the ask. E.g. do you just want to see the percentage? Make a dataset? Or merge the percentage into the data? If you just want to see them, they can be obtained with:

bysort did: tabulate male

If it is other things that is needed, then please clarify.

[smith Jason]

I want to create a new variable--male_p, which means the percentage of male by distrcit.
Thanks!

[Ken Chui]

I want to create a new variable--male_p, which means the percentage of male by distrcit.
Thanks!

Since 1 is male and 0 is female, the mean of it would be the proportion of male in fraction.

egen prop_male = mean(male), by(did)

If you need this in percent, then just follow with a generate to create a new one multiplied by 100.

[smith Jason]

Since 1 is male and 0 is female, the mean of it would be the proportion of male in fraction.

egen prop_male = mean(male), by(did)

If you need this in percent, then just follow with a generate to create a new one multiplied by 100.

In fact, this is the long format data. For district 1, male_prop=3/7=42.85%

[smith Jason]

My code is as follows and it can work now,

egen tag_id = tag(id)
egen NUM = total(tag_id), by(did)

egen tagid = tag(id) if male==1
egen num = total(tagid), by(did)

gen male_prop=num/NUM

LOL. It is a good opportunity for me to learn, grow, and think over and over in practice.

[Ken Chui]

My code is as follows and it can work now,

egen tag_id = tag(id)
egen NUM = total(tag_id), by(did)

egen tagid = tag(id) if male==1
egen num = total(tagid), by(did)

gen male_prop=num/NUM

LOL. It is a good opportunity for me to learn, grow, and think over and over in practice.

This is sufficient, no need for another tag:

egen tag_id = tag(id)
egen NUM = total(tag_id), by(did)
egen num = total(tag_id * male), by(did)
gen male_prop=num/NUM

[smith Jason]

Thank you!