generating an accumulating column within groups

Ziba — Mon, 28 Jan 2019 00:56:15 GMT

Dear All,

my question might be too simple but I actually have some trouble figuring it out...I hope I can get some help from you guys, thanks in advance.

 data count;
 set patent_list;
  by permno;
  if first.permno=1 then NumCit=0;
  NumCit+ncites;
  run;

I used the above code to calculated the accumulating number of citation for each company(see the table below). But what I want is only the total number of the citation. (that means I dont need other rows indicating the current accumulated values)

Now I want to create two columns, one is the total number of the patents for each company (permno is the unique identifier, so you can count it until the permno change to different permno, ) and the total number of citations that each company received for its all applied patents.

for example, company D has 10 patents(count till permno changes) and 28 citation in total. So, as a final result, I want to keep all the variables from the table, and add these two new columns.

name	permno	merged_file_date	patnum	fdate	ncites
A	11342	2/9/2004
B	11760	3/17/2003
C	12320	11/30/2004
D	44150	10/20/2004	6412747	3/22/2001	9
D	44150	10/20/2004	6412746	3/22/2001	5
D	44150	10/20/2004	6468373	4/20/2000	3
D	44150	10/20/2004	6405989	3/22/2001	2
D	44150	10/20/2004	6405990	3/22/2001	3
D	44150	10/20/2004	6554243	3/22/2001	6
D	44150	10/20/2004	6905396	11/20/2003
D	44150	10/20/2004	6874802	2/25/2003
D	44150	10/20/2004	7000935	12/22/2003
D	44150	10/20/2004	6949162	9/23/2002
E	47271	4/1/2003
F	55060	5/20/2004	6550287	6/7/2002	0
G	65496	6/23/2004	6537459	4/20/2000	11
G	65496	6/23/2004	6562249	7/10/2001
G	65496	6/23/2004	6682875	6/18/2001
G	65496	6/23/2004	6365057	11/1/1999	3
G	65496	6/23/2004	6468439	11/1/1999	9
G	65496	6/23/2004	6409930	11/1/1999	0
F	65964	4/8/2003	5952619	10/30/1998	8
F	65964	4/8/2003	5980057	9/15/1998	11
F	65964	4/8/2003	6204824	9/22/1998	4
F	65964	4/8/2003	6211844	1/7/1999	3
H	67432	3/31/2004	6572805	10/13/1999	0
H	67432	3/31/2004	6656230	10/11/2001	0
H	67432	3/31/2004	6686037	12/16/1999	1
H	67432	3/31/2004	6322654	10/13/1999	3
H	67432	3/31/2004	6839937	9/25/2002
I	75597	5/19/2003	6679273	7/23/2001
I	75597	5/19/2003	6628378	11/9/1999
I	75597	5/19/2003	6264757	7/23/1998	4
I	75597	5/19/2003	6280596	10/1/1999	2
I	75597	5/19/2003	6035689	6/24/1998	3

Re: generating an accumulating column within groups

Tom — Mon, 28 Jan 2019 01:20:58 GMT

Use the LAST.PERMNO flag to tell which observation to output.

To count the observations just use another counter variable that you increment by 1 each time.

data count;
  set patent_list;
  by permno;
  if first.permno=1 then call missing(NumPat,NumCit);
  NumCit+ncites+0;
  NumPat+1;
  if last.permno;
run;

Re: generating an accumulating column within groups

Ziba — Thu, 31 Jan 2019 15:11:43 GMT




data logit.count;
  set logit.patent_company;
  by permno;
  if first.permno then Numcit = 0;
  	numcit + ncites;
  if first.permno then 
  	Numpat = 0;
  	if patnum ne '' then Numpat + 1;
	else Numpat + 0;
	
  if last.permno;
run;

hey, thanks for your code. Just made several changes.

I used the above code as I dont want to count patent number if the company doesnt have any patents.

topic generating an accumulating column within groups in SAS Programming

generating an accumulating column within groups

Re: generating an accumulating column within groups

Re: generating an accumulating column within groups