topic Segregate the dataset based on unique values of a variable in New SAS User

Segregate the dataset based on unique values of a variable

chandan_mishra — Wed, 21 Aug 2019 16:27:57 GMT

Hello

I want to segregate the dataset based on the unique values of a variable:

Ex. The database name is DB. Date is one of the column in MONYY5. format:

Date

Oct14

Nov14

Dec14

Jan15

Feb15

Mar15

Apr15

data OCT14;
set DB;
where MONTH(Date) EQ 10 and YEAR(Date) Eq 2014;
run;

data NOV14;
set DB;
where MONTH(Date) EQ 11 and YEAR(Date) Eq 2014;
run;

data DEC14;
set DB;
where MONTH(Date) EQ 12 and YEAR(Date) Eq 2014;
run;

This has to be repeated number of times as the unique values in the date column.

The general logic which I am trying to get is:

Data Oct14;

set DB;

where date=Oct14;

run;

Just write this one and loop it for all the unique values available in the date column.

Thanks

Chandan Mishra

Re: Segregate the dataset based on unique values of a variable

novinosrin — Wed, 21 Aug 2019 17:04:02 GMT

Hello @chandan_mishra Please refer to this link , use any example and replicate to your needs. Very well detailed there

http://www.sascommunity.org/wiki/Split_Data_into_Subsets

And another one

https://blogs.sas.com/content/sasdummy/2015/01/26/how-to-split-one-data-set-into-many/

http://sasnrd.com/sas-split-dataset-by-group/

Re: Segregate the dataset based on unique values of a variable

Reeza — Wed, 21 Aug 2019 17:06:39 GMT

Please read the first part - which is this is usually not needed or recommended and BY group processing is very efficient and simple to program. If after this you find yourself creating many macro loops you know that splitting your data is a bad idea. There are several use cases when this is required - splitting data that is too big for faster processing but that usually means you'll have tens of millions of rows of data. Another usage can be accomplished in other manners that are more efficient, ie a dynamic WHERE statement for PROCs is one option.

@chandan_mishra wrote:

Hello

I want to segregate the dataset based on the unique values of a variable:

Ex. The database name is DB. Date is one of the column in MONYY5. format:

Date

Oct14

Nov14

Dec14

Jan15

Feb15

Mar15

Apr15
data OCT14;
set DB;
where MONTH(Date) EQ 10 and YEAR(Date) Eq 2014;
run;

data NOV14;
set DB;
where MONTH(Date) EQ 11 and YEAR(Date) Eq 2014;
run;

data DEC14;
set DB;
where MONTH(Date) EQ 12 and YEAR(Date) Eq 2014;
run;
This has to be repeated number of times as the unique values in the date column.

The general logic which I am trying to get is:

Data Oct14;

set DB;

where date=Oct14;

run;

Just write this one and loop it for all the unique values available in the date column.

Thanks

Chandan Mishra