Numbering with in groups and missing data

[Solvej]

Hi Again,

I always get great help from this Group.

I have a dataset where I want to number with in Groups. I have done this before with out problems however this time is complicated for some reason I cannot understand.

I have

{

data have; input Record_id date_of_admission ddmmyy10. admssion var1 $ var2 $ var3 $ var4 $ date_of_ques ddmmyy10. var5 $ ;

datalines;

1 07-01-2010 1 A B D . . .

1 09-01-2010 2 F D G X 10-01-2010 B

1 09-02-2010 3 A D G X 09-03-2010 A

1 09-02-2010 3 A D G X 12-03-2010 A

1 09-02-2010 3 A D G X 14-03-2010 A

1 09-03-2010 4 F D G . . .

1 09-03-2017 5 X T V . . .

2 07-02-2009 1 C R V X 14-07-2009 C

2 03-09-2010 2 C T V . . .

;

run;

What I want is number with in Groups defined by record_id date of admission, but only for the lines where var4=x.

　

datalines;

1 07-01-2010 1 A B D . . .

1 09-01-2010 2 F D G X 10-01-2010 B 1

1 09-02-2010 3 A D G X 09-03-2010 A 1

1 09-02-2010 3 A D G X 12-03-2010 A 2

1 09-02-2010 3 A D G X 14-03-2010 A 3

1 09-03-2010 4 F D G . . .

1 09-03-2017 5 X T V . . .

2 07-02-2009 1 C R V X 14-07-2009 C 1

2 03-09-2010 2 C T V . . .

;

run; }

I have tried the following however this only produces numbers with record_id.

{ 

data WANT;

set have;

NUMMERDRUGc + 1;

by record_id date_of_admission var4;

if first.record_id and first.date_of_admission and first.var4 then NUMMERDRUGc = 1;

run;

 }

Any suggestions.

Kind regards

Solvej

[RW9]

Just split the data, add the count for data where var4=X, then add back to other data and sort:

data have; 
  input record_id date_of_admission ddmmyy10. admission var1 $ var2 $ var3 $ var4 $ date_of_ques ddmmyy10. var5 $;
datalines;
1 07-01-2010 1 A B D . . .
1 09-01-2010 2 F D G X 10-01-2010 B
1 09-02-2010 3 A D G X 09-03-2010 A
1 09-02-2010 3 A D G X 12-03-2010 A
1 09-02-2010 3 A D G X 14-03-2010 A
1 09-03-2010 4 F D G . . .
1 09-03-2017 5 X T V . . .
2 07-02-2009 1 C R V X 14-07-2009 C
2 03-09-2010 2 C T V . . .
;
run;
data x1;
  set have (where=(var4="X"));
  retain cnt 0;
  by record_id date_of_admission admission;
  if first.admission then cnt=cnt+1;
run;

data want;
  set have (where=(var4 ne "X")) x1;
run; 

Please note how I use the code window - its the {i} above post area - to post code.

[Solvej]

HI Again

Thank you for your answer.

When I run the code I get this result.

Obs Record_id date_of_admission admission var1 var2 var3 var4 date_of_ques var5 cnt123456789

1	18269	1	A	B	D		.		.
1	18271	2	F	D	G	X	18272	B	1
1	18302	3	A	D	G	X	18330	A	2
1	18302	3	A	D	G	X	18333	A	2
1	18302	3	A	D	G	X	18335	A	2
1	18330	4	F	D	G		.		.
1	20887	5	X	T	V		.		.
2	17935	1	C	R	V	X	18092	C	3
2	18508	2	C	T	V		.		.

This is not exactly what I had in  mind. However I have tried to exchange the admission variable with a multiple of variables but still not getting it right.

What am I missing?

Solvej

[RW9]

Sorry, I can't really tell what it is you want:

1 09-01-2010 2 F D G X 10-01-2010 B 1

1 09-02-2010 3 A D G X 09-03-2010 A 1

1 09-02-2010 3 A D G X 12-03-2010 A 2

Why does it only change on row 3?  Sounds like lagged result or something, i.e. if the previous result is the first one then add 1.

[Solvej]

Hi Again

The reason for the change in row three is that row two and three two different admissions. Thus I want to number the questionnaires based on the dates of questionnaires within each admission.

Kind regards

Solvej

[Ksharp]

data have; 
  input record_id date_of_admission : $20. admission var1 $ var2 $ var3 $ var4 $ date_of_ques : $20. var5 $;
datalines;
1 07-01-2010 1 A B D . . .
1 09-01-2010 2 F D G X 10-01-2010 B
1 09-02-2010 3 A D G X 09-03-2010 A
1 09-02-2010 3 A D G X 12-03-2010 A
1 09-02-2010 3 A D G X 14-03-2010 A
1 09-03-2010 4 F D G . . .
1 09-03-2017 5 X T V . . .
2 07-02-2009 1 C R V X 14-07-2009 C
2 03-09-2010 2 C T V . . .
;
run;
data want;
 set have;
 by record_id var5 notsorted;
 if first.var5 then n=0;
 n+1;
 if missing(var5) then call missing(n);
 run;
 
 proc print noobs;run;

[Solvej]

Thank you Ksharp.

I see that you programme Works, but I see that my example is bad. I apologize for this.

I also needs to work when my data looks like this

{i}

data have; input Record_id date_of_admission ddmmyy10. admission var1 $ var2 $ var3 $ var4 $ date_of_ques ddmmyy10. var5 $ ;

datalines;

1 07-01-2010 1 A B D . . .

1 09-01-2010 2 F D G X 10-01-2010 B

1 09-02-2010 3 A D G X 09-03-2010 A

1 09-02-2010 3 A D G X 12-03-2010 G

1 09-02-2010 3 A D G X 14-03-2010 B

1 09-03-2010 4 F D G . . .

1 09-03-2017 5 X T V . . .

2 07-02-2009 1 C R V X 14-07-2009 C

2 03-09-2010 2 C T V . . .

;

run;

[Solvej]

I changed your code to this and now it seems to work. Thank you.

{i}

data want;

set have;

by record_id admission notsorted;

if first.admission then n=0;

n+1;

if missing(var5) then call missing(n);

run;

proc print noobs;run;