replacing value of another row from the current row

[Henry Han WLG]

serial	numprec	pernum	fchloc	momloc	firstchild	nchild	nchild2	age
1184821	18	1		0		9+		53
1184821	18	2		0		9+	9	52
1184821	18	3	3	2	1	1 child	9	27
1184821	18	4		2		1 child	9	26
1184821	18	5		2		1 child	9	23
1184821	18	6		2		0 childr	9	21
1184821	18	7		2		1 child	9	19
1184821	18	8		2		0 childr	9	16
1184821	18	9		2		0 childr	9	14
1184821	18	10		2		0 childr	9	12
1184821	18	11		2		0 childr	9	10
1184821	18	12		0		1 child		27
1184821	18	13		0		1 child		23
1184821	18	14		0		1 child		21
1184821	18	15	15	12	1	0 childr	1	5
1184821	18	16	16	13	1	0 childr	1	3
1184821	18	17	17	14	1	0 childr	1	less tha
1184821	18	18	18	7	1	0 childr	1	less tha

existing variable nchild has a limit of 9 but there are data where the actual value is higher (this specific instance is not).
I've created 3 new variables at an attempt to somehow add firstchildlocation to the mother

by bysort serial momloc: gen nchild2 = _N if momloc != 0
gen firstchild = 1 if eldch[_n-pernum+momloc] == age & age < 18
gen fchloc = pernum if firstchild == 1

I would like to do, for example, for pernum3

replace momloc[_n-pernum+momloc] = momloc[_n] if firstchild == 1

I understand that I can only do this the other way around.

I've been searching on google for 'copying (and other similar words) values to different rows on stata'.

Any kind of command or other "search terms" that are more in line with what I'm looking for would be much appreciated.

[Phil Bromiley]

You didn't get a quick answer. You'll increase your chances of a useful answer by following the FAQ on asking questions - provide Stata code in code delimiters, readable Stata output, and sample data using dataex. Many will be happier if you use Stata terms - we have observations, not rows. It would also help if your example were very carefully tailored to show your problem. Only include the variables you need. Also, I don't think you have a firstchild==1 in your example. Without that, we can't duplicate your problem. It would also help if you explain the structure of your data more. I don't know what all the identifiers are.

I have trouble understanding firstchildlocation since it doesn't appear in your data sample. I'm not sure what you're doing, but a couple of observations. You probably want to change from strings like 10 childr to a variable equal 10. This can be done with the string functions that identify the first space and then another that copies from the beginning to that space. Then convert the number to a real.

If everything is real, you can do by with egen and max to set values to the max for a group.