linear interpolation

[AmirSari]

Hi all,

I am trying to add US treasury yield to my dataset based on issue and maturity dates. For observations without exact matches I want to use linear interpolation. Matching the observation with the same issue and maturity dates is easy, but using linear interpolation for those that don't have a exact maturity match is what I am having trouble with. Here is samples of my dataset. I have also attached the US treasury yield dataset.

MySample:

Issuedate     MaturityDate     ticker

1/11/1990    1/15/1994           EK

1/18/1990    1/15/1996           CO

3/07/1990    3/15/2002           DP

There are multiple observations in the treasury file that have the same issue date but different maturity dates. To do linear interpolation, for any observation in my data, I need the two closest observations on maturity dates in the treasury file so that I could use proc expand and linear interpolate. If I could get something like this, then I can use proc expand and do linear interpolation.

Issuedate     MaturityDate    ticker          Issue              Maturity              Yield

1/11/1990    1/15/1994           EK        1/11/1990          1/11/1993            7.95

1/11/1990    1/15/1994           EK        1/11/1990          1/11/1994               .

1/11/1990    1/15/1994           EK        1/11/1990          1/11/1995            7.94

1/18/1990    1/15/1996           CO       1/18/1990          1/18/1995            8.27

1/18/1990    1/15/1996           CO       1/18/1990          1/18/1996               .

1/18/1990    1/15/1996           CO       1/18/1990          1/18/1997            8.31

3/07/1990    3/15/2002           DP       3/07/1990          3/07/2000            8.58

3/07/1990    3/15/2002           DP       3/07/1990          3/07/2002               .

3/07/1990    3/15/2002           DP       3/07/1990          3/07/2010               .

Any help is much appreciated

[johnsville]

data samples ;
infile datalines ;
input
    issuedate :mmddyy10.
    maturitydate :mmddyy10.
    ticker :$2.
;
format issuedate maturitydate mmddyys10. ;
datalines;
1/11/1990    1/15/1994           EK
1/18/1990    1/15/1996           CO
3/07/1990    3/15/2002           DP
;

libname UST "/folders/myfolders/data/treasuries" ;

proc sql noprint ;
    create table two as
    select s.issuedate
        , s.maturitydate
        , s.ticker
        , t.issue
        , t.maturity
        , t.yield
    FROM samples s
    JOIN UST.treasury t
    on s.issuedate eq t.issue
    and s.maturitydate le t.maturity
    ;
quit ;