Using Arrays to sum obs like proc sql sum function

[Arum]

Hello I am trying to sum a large amount of obs called NAICS4 which have 390 distinct values. The taxable??? colume has 90 distinct columes and will increase monthly. I can do a sum statement in proc sql like this:

proc sql; create table state4SortSum as select

naics4, sum(taxable190) as sumTaxable190, sum(taxable191) as sumTaxable191, sum(taxable390) as sumTaxable390

from rg.state4sort

group by naics4; quit;

But I am not wanting to write 90+ sum statements, so I tried to write a array:

data stateTest; length item $4;

  array txSum{*} taxable190-taxable211 ;

do i = 1 to dim(txSum);

  set rg.state4sort;

  by naics4;

  item =naics4;

  tx = sum( of txSum{i})  ;

  if last.naics4 then output;

end;

drop tx;

run;

However this array is only giving me the value of the last obs it runs through for each NAICS4 and I want it to give me the sum of all like it would in SQL. Any help with a better way to get the equivialent of the SQL without writing out everything would be great.  

 

naics4	taxable190	taxable290	taxable390	taxable490	taxable191	taxable291	taxable391	taxable491	taxable192	taxable292	taxable392	taxable492
1111	-0.001	-0.001	-0.001	-0.001	-0.001	-0.001	-0.001	-0.001	-0.001	-0.001	-0.001	-0.001
1111
1111
1112	-0.001	-0.001	-0.001	-0.001	-0.001	-0.001		-0.001	-0.001	-0.001	-0.001	-0.001
1112	-0.001	-0.001		-0.001	-0.001	-0.001	-0.001			-0.001	-0.001	-0.001
1112
1112
1113	23398	15296	21359	64786	23673	15961	17453	159439	27648	24140	-0.001	236359
1113	-0.001	-0.001	-0.001	-0.001	-0.001	-0.001	-0.001	-0.001	-0.001	-0.001		-0.001
1113		-0.001						-0.001				-0.001
1113										-0.001	-0.001	-0.001
1113				-0.001			-0.001	131411	-0.001	-0.001	-0.001	4772
1113
1113	-0.001	-0.001	-0.001	-0.001	-0.001	-0.001	-0.001	-0.001	-0.001	-0.001	-0.001	-0.001

[Tom · Accepted solution]

The ARRAY statement is to allow you to work across mulitple variables in the same observation.

The SUM() function in PROC SQL is to allow you to work across multiple obsevations.

You want to use PROC SUMMARY to sum your variables.

proc summary data=rg.state4sort nway ;

  class naics4;

  var Taxable: ;

  output out=state4SortSum sum=;

run;

You can work out how to rename the variables from Taxable ... to SumTaxable...

[Linlin]

anothe approach:

example:

data have;

input naics4 $ taxable1-taxable5;

cards;

1 20 30 40 50 60

1 25 35 45 55 65

1 30 40 50 60 70

2 20 30 40 50 60

2 25 35 45 55 65

3 30 40 50 60 70

;

run;

proc sql noprint;

select catt('sum','(',name,')',' as ',catt(' sum',name))into : names separated by ','

  from dictionary.columns

    where libname='WORK' and memname='HAVE' and upcase(name) like 'TAXABLE%' ;

quit;

%put &names;

proc sql;

  create table want as select naics4,&names

    from have

       group by naics4;

quit;

Linlin

[Haikuo]

An array approach, borrowed examples from LinLin's post:

data have;

input naics4 $ taxable1-taxable5;

cards;

1 20 30 40 50 60

1 25 35 45 55 65

1 30 40 50 60 70

2 20 30 40 50 60

2 25 35 45 55 65

3 30 40 50 60 70

;

run;

data want (drop=tax:);

do until (last.naics4);

set have;

by naics4;

array tax(*) taxable1-taxable5;

array st(*) sum_tax1-sum_tax5;

do _n_=1 to dim(tax);

   st(_n_)+tax(_n_);

end;

end;

output;

call missing (of _all_);

run;

proc print;run;

Regards,

Haikuo