01-30-2014 02:40 AM
I need to read one month previous / last month data from one folder having historical data. like In February I need to read January data. my data sets having below format. I need to automate the code. pls suggest me if any code available.
01-30-2014 02:52 AM
By format, I assume you mean data set name naming standard.
For 2), just create a macro variable that corresponds to the YYYYMM part, by using the INTNX function.
For 1), are there a data set for each day in the previous month? How does your code look like that consume these data sets?
In this scenario, my guess that you need some kind of macro that loops over the input data set names, based on the dates in the previous month.
Either way, it could be useful to have logic that lets you input the month of choice, if a situation occurs when you need to reload data older than the previous month.
01-30-2014 05:08 AM
I receive data on monthly. Every month my data set name is same but extension month and year(if old year) is change in some data sets having date extension. I want to take only previous month data set based on month and year. if any code pls suggest me.
02-02-2014 04:58 AM
You did not specify whether the input tables were in SAS format or external files that need to be imported into SAS.
If the tables are already SAS datasets you may in the same library then you can get a lis of all SAS datasets in the library using SQL, something like this (untested)
Proc SQL ;
Create table monthly_tables as
, Scan (MEMNAME, -1, '_') As Date_Suffix
Where Library = 'HAVE' /* Library name must be in caps */
Having missing (compress (Date_Suffix, ' ', 'D') ) /* suffix must be all numerals */
%Let LastMonth = %Sysfunc (INTNX (Month, %Sysfunc (Today ()), -1)) ;
Data Wanted_Tables ;
Set Monthly_Tables ;
Retain Last_Month &LastMonth ;
If Length (Date_Suffix) = 8 then
If INTNX ('MONTH', Input (Date_Suffix, YYMMDDN8.), 0) = last_Month
then Output ;
Else If Length (Date_Suffix) = 6 then
If Input (Date_Suffix, YYMMN6.) = last_Month
then Output ;
Else Delete ;
That should give you a list of files to use. You do not specify what the next step of your process should be but it could be put into a macro and then invoked using Call Execute () in a data _Null_ step.
If your data are in external files this method could be adapted but you would need to use a method to read the directory / directories containing the files, and use Scan (filename, -2, '._') in the first step (note the additional period).
02-02-2014 05:24 AM
As Linus said: Populate a macro variable which you then can use in the dynamic part of the name.
%put yymmdd_begin_prev_month= &yymmdd_begin_prev_month;
%put yymm_prev_month= &yymm_prev_month;
02-02-2014 11:41 AM
To combine a bunch of datasets with similar names try using the : modifier in the SET statement. Here is an example.
* Make some datasets for previous and current month for testing ;
data xyz_201401 xyz_20140101 xyz_20140105 xyz_20140201 ;
* Get basename for previous month in XYZ_yyyymm format ;
* Combine all datasets for previous month ;
data want ;
length ds dsname $41 ;
set &base: indsname=ds ;
dsname = ds;
put _n_ ds= ;