concatenate week(sasdate) in a string and retain leading zero for weeks 01 to 09

[yelkenli]

i am having trouble with a simple format need.  

My first attempt (which was with CATS and no PUT statement) resulted in single digit weeks for weeks 1 through 9.  

The below results in a single digit week number with a blank space where a zero should be.  The rest of data manipulations work but I would prefer to have the zero.  

The remaining data manipulations provide a table showing a volume over a time line by week (across years, etc.).  a second path might be to skip creating this variable and present the data some other way but I have not see that alternate path.    so help with keeping the zero is appreciated. 

create table want AS
SELECT 
      some variables
      ,CAT('Yr', year(datepart(a_date)) , 'FW', put(week(datepart(a_date),'v'), 2.) ) AS FW_filter

from atable; 

results are as such:

Yr2017FW 1

Yr2017FW 2

etc.

would like 

Yr2017FW01

Yr2017FW02

[Reeza · Accepted solution]

Change 2. to Z2, which keeps the leading 0's. 

CAT('Yr', year(datepart(a_date)) , 'FW', put(week(datepart(a_date),'v'), Z2.) ) 

Or use a custom format instead.

Create format:

proc format;
picture myCustomFmt (default=20)  low-high = 'Yr%YFW%0U' (datatype=datetime ) ;
run;

use format:

put(a_date, myCustomFmt.) as FW_filter

---

@yelkenli wrote:

i am having trouble with a simple format need.  

My first attempt (which was with CATS and no PUT statement) resulted in single digit weeks for weeks 1 through 9.  

The below results in a single digit week number with a blank space where a zero should be.  The rest of data manipulations work but I would prefer to have the zero.  

The remaining data manipulations provide a table showing a volume over a time line by week (across years, etc.).  a second path might be to skip creating this variable and present the data some other way but I have not see that alternate path.    so help with keeping the zero is appreciated. 

create table want AS
SELECT 
      some variables
      ,CAT('Yr', year(datepart(a_date)) , 'FW', put(week(datepart(a_date),'v'), 2.) ) AS FW_filter

from atable; 

results are as such:

Yr2017FW 1

Yr2017FW 2

etc.

would like 

Yr2017FW01

Yr2017FW02

---

[ballardw]

A custom format may be the easiest way, IMHO:

proc format library=work;
picture mydtfmt (default=9)
low-high = 'Y%YFW%0V' (datatype=datetime)
;
run;

data work.example;
   dt="15JAN2019:12:24:00"dt;
   y = put (dt,mydtfmt.);
run;

You can use either the %U %V or %W week definition depending which you want. The 0 in %0V places a 0 when there is only a single digit to the week. By declaring the format for use with datetimes you need not extract the date part. The CASE of the % directives is case sensitive. Use of %y would get a 2 digit year. The first Y and FW are literal strings inserted.