indsname adding extra datasets

Steelers_In_DC — Tue, 14 Nov 2017 21:31:09 GMT

Hello all. I have some code that I run on a monthly basis, I am using indsname to identify sources and initialize new variables.

The code is here:

data fyf_output_graph_append;
set
   average_lt700 : indsname = set
    average_700_800 : indsname = set
   average_gt800 : indsname = set
   average_bus : indsname = set;
source1 = set;
length Fico $7.;
if source1 = 'WORK.AVERAGE_LT700' then do;
   asset_class = 'Consumer';
   Fico = '< 700';
end;
else if source1 = 'WORK.AVERAGE_700_800' then do;
   asset_class = 'Consumer';
   Fico = '700-800';
end;
else if source1 = 'WORK.AVERAGE_GT800' then do;
   asset_class = 'Consumer';
   Fico = '> 800';
end;
else if source1 = 'WORK.AVERAGE_BUS' then do;
   asset_class = 'Business';
   Fico = '';
end;
else delete;
drop source1;
Date = intnx('month',input(month,yymmn6.),0,'e');
format date mmddyy10.;
drop month;
run;

The reason I put the 'else delete' is because there are temporary datasets generated through this process, here is the log:

NOTE: There were 456 observations read from the data set WORK.AVERAGE_LT700.
NOTE: There were 480 observations read from the data set WORK.AVERAGE_LT700_PREP.
NOTE: There were 480 observations read from the data set WORK.AVERAGE_LT700_PREP_TWO.
NOTE: There were 456 observations read from the data set WORK.AVERAGE_700_800.
NOTE: There were 480 observations read from the data set WORK.AVERAGE_700_800_PREP.
NOTE: There were 480 observations read from the data set WORK.AVERAGE_700_800_PREP_TWO.
NOTE: There were 456 observations read from the data set WORK.AVERAGE_GT800.
NOTE: There were 480 observations read from the data set WORK.AVERAGE_GT800_PREP.
NOTE: There were 480 observations read from the data set WORK.AVERAGE_GT800_PREP_TWO.
NOTE: There were 456 observations read from the data set WORK.AVERAGE_BUS.
NOTE: There were 480 observations read from the data set WORK.AVERAGE_BUS_PREP.
NOTE: There were 480 observations read from the data set WORK.AVERAGE_BUS_PREP_TWO.

I do not understand why the prep and prep_two datasets are there, but I don't delete them my dataset grows in triplicate.

I tried to duplicate the issue and I am not able, I am pasting the log for a similar piece of code below, where the temp datasets are not generated:

24         GOPTIONS ACCESSIBLE;
25         data join_all;
26         set fico_lt700 : indsname = source
27             fico_700_800 : indsname = source
28             fico_800_plus : indsname = source
29             business : indsname = source;
30         name=source;
31         length fico $7.;
32         if name = 'WORK.FICO_700_800' then do;
33         fico = '700-800';
34         asset_class = 'Consumer';
35         end;
36         else if name = 'WORK.FICO_800_PLUS' then do;
37         fico = '800';
38         asset_class = 'Consumer';
39         end;
40         else if name = 'WORK.FICO_LT700' then do;
41         fico = '< 700';
42         asset_class = 'Consumer';
43         end;
44         else if name = 'WORK.BUSINESS' then do;
45         fico = '';
46         asset_class = 'Business';
47         end;
48         Date = load_date;
49         drop name;
50         run;

NOTE: There were 3080798 observations read from the data set WORK.FICO_LT700.
NOTE: There were 2914217 observations read from the data set WORK.FICO_700_800.
NOTE: There were 1716779 observations read from the data set WORK.FICO_800_PLUS.
NOTE: There were 1643479 observations read from the data set WORK.BUSINESS.
NOTE: The data set WORK.JOIN_ALL has 9355273 observations and 6 variables.

Can someone explain why the temp datasets are generated with one and not the other?

Thank You in advance.

Re: indsname adding extra datasets

Reeza — Tue, 14 Nov 2017 21:48:37 GMT

What are you expecting the colon to do in your code?

It's likely being interpreted as being part of the data set name and means include all datasets that start with that prefix - which is hwy you're getting the extra data sets.

You also don't need multiple INDSNAME options, a single one is all that's required.

I think this may be what you want?


set 
    average_lt700 
    average_700_800 
    average_gt800
    average_bus indsname = set;

Re: indsname adding extra datasets

Steelers_In_DC — Tue, 14 Nov 2017 22:13:07 GMT

I thought the : was part of the syntax, I saw it somewhere and thought it was necessary.

Now that you say that it makes so much sense. I'll update my code accordingly.

Thank You,

Re: indsname adding extra datasets

Reeza — Tue, 14 Nov 2017 22:14:17 GMT

You should also use a different word than SET, it's likely to lead to confusion since it's used in SAS for as a STATEMENT. I don't think you'll have any technical issues, but it's not recommended.

topic Re: indsname adding extra datasets in SAS Programming

indsname adding extra datasets

Re: indsname adding extra datasets

Re: indsname adding extra datasets

Re: indsname adding extra datasets