topic Re: Different date format to single format in SAS Programming

Different date format to single format

Sathish_jammy — Wed, 11 Nov 2020 07:38:24 GMT

Hi,

I need to convert the texted date variable to a particular format.

Eg:

data have;
input R_Date$20.;
cards;
31/01/2000
15/3/2015
30-5-2010
30-12-2018
31/7/20
15/2/12
18/05/15
11-08-08
2000
1998
;

Expected result

01/31/2000

03/15/2015

05/30/2010

12/30/2018

07/31/2020

02/15/2012

05/18/2015

08/11/2008

01/01/2000

01/01/1998

Kindly suggest a code to get MM/DD/YYYY format.

Re: Different date format to single format

Kurt_Bremser — Wed, 11 Nov 2020 08:29:54 GMT

This is not data, this is unusable crap.

11-08-08

could be

2011-08-08

2008-08-11

2008-11-08

not even the year can be determined with any confidence. So there simply is no way to reliably make dates out of this. And reliable processes are what we need and strive for. To roll dice, I don't need a data warehouse.

Anybody who still uses 2-digit years after Y2K is * ***** ***** in my eyes, and needs to be cared for in a proper institution for the mentally challenged.

Print that "data" on a sheet, wrap it around a baseball bat, and go to the person that sent you this. Apply the baseball bat liberally. Maybe some neurons get into contact and start to fire.

Sorry for my rant, but things like this severely damage my belief in a long-term future for the human species.

PS anybody who finds exaggeration or irony can keep it and use it freely.

Re: Different date format to single format

andreas_lds — Wed, 11 Nov 2020 08:32:54 GMT

EDIT: i removed my "solution" after reading the fully valid comment by @Kurt_Bremser

Re: Different date format to single format

Tom — Wed, 11 Nov 2020 16:14:31 GMT

So are you assuming the values should be in DMY order? If so use the DDMMYY informat. If you want to get the year only values you will need to add more logic to handle those cases.

I wouldn't compound the confusing by using either DMY or MDY order for displaying the derived dates. Use an unambiguous format like YYMMDD10. or DATE9. .

data want;
  set have ;
  date=coalesce(input(r_date,??ddmmyy10.),input(r_date,??anydtdte20.));
  if missing(date) and length(r_date)=4 and not missing(input(r_date,??4.))
    then date=mdy(1,1,input(r_date,4.))
  ;
  format date yymmdd10.;
run;

proc print;
run;

Obs    R_Date              date

  1    31/01/2000    2000-01-31
  2    15/3/2015     2015-03-15
  3    30-5-2010     2010-05-30
  4    30-12-2018    2018-12-30
  5    31/7/20       2020-07-31
  6    15/2/12       2012-02-15
  7    18/05/15      2015-05-18
  8    11-08-08      2008-08-11
  9    2000          2000-01-01
 10    1998          1998-01-01