topic Re: merge using month and year ? in ODS and Base Reporting

merge using month and year ?

pkmkart — Fri, 03 Apr 2015 16:44:09 GMT

Hi,

I have month and year variable which i extracted from date variable, now i need o merge to other dataset which has date, but i want to merger only by their months and year not on their exact date?

how can i do that? If i try to extract year and month and try to merge by year and month, some times the results are correct, for example if it is one year worth data then i have no issue, when the dates are between two years then the problem comes, it throws me error that month variable is not properly sorted? Anyone Please help me

Re: merge using month and year ?

Reeza — Fri, 03 Apr 2015 17:12:46 GMT

1. Create a date variable using MDY() function. i.e. new_date_var=mdy(month_var, 1, year_var)

2. Use a SQL join where you use a PUT() on the join condition i.e. put(new_date_var, monyy7.)=put(other_date, monyy7.)

Re: merge using month and year ?

ballardw — Fri, 03 Apr 2015 18:29:15 GMT

It would help to provide example code of what you have attempted and a few records from each data set that demonstrate your difficultly.

Any time you use a data step MERGE with a BY statement all data sets must be sorted by those variables.

You probably need to do something like this before the merge:

Proc sort data=dataset1; by year month;run;

Proc sort data=dataset2; by year month;run;

data merged;

merge dataset1 dataset2;

by year month;

run;

Re: merge using month and year ?

Ksharp — Sat, 04 Apr 2015 08:36:31 GMT

Use a special option of BY.

data merged;

merge dataset1 dataset2;

by date groupformat;

format date monyy7.;

run;

Xia Keshan