Hi,  Is there a way SAS allows to do a mean across varaibles (and use a BY statement). Here's what I mean: HAVE: Prod   Price1  Price2  Price3  Price4 A          2          4          7         6 B          3          .           5        3 C          .          5          .         . D          .           .           .         . WANT:   If I want the MEAN of the last 3 Prices & last 4 Prices Prod   Month3_Prices         Month4_Prices A                5.6                            4.7 B                 4                              3.6 C                 5                                5 D                 .                                .       I tried different variations of Proc MEAN, couldn't get it, but I'm sure there is a simple way. Thanks ... View more

This is definately not an easy task.. so total kudos if anyone that can answer this. My task is to merge 2 tables into a main 1 (so 3 all together). One of the 2 has data that is aggregated by city (so population by city) and the 2nd is aggregated by regions not in a city in table 1 (so out of city population).... The 3rd main one has data not aggregated, it's at the Postal Code level, but attached to each row is a City Name and an Out-of-City name.  So far so good.  When I merge all 3, the postal Codes that belong in the City (eg. Toronto) will get Toronto's total population, and that would be double/triple/etc. counting, considering Toronto has many many postal codes.  So my task is to equally divide Toronto's population into each of the many many Postal Codes residing under Toronto in List 3.  All this by different dates as well.. Confused ?, let me try visually..   THANK YOU.. Table 1 CITY             Date                   Pop Toronto         2011.01           5,795,389 Toronto         2011.02           5,803,064 Toronto         2011.03           5,810,739 Table2 (Note, Ontario is non-Toronto) CITY            Date                   Pop Ontario         2011.01           2,678,462 Ontario         2011.02           2,678,754 Table3 POSTALCODE        City                   Date   M1P 3F7                 Toronto             2011.01 M1P 3F7                 Toronto             2011.02       M1J 2G8                  Toronto            2011.01 M1J 2G8                  Toronto            2011.02 M1J 2G8                  Toronto            2011.03          L0A 2F6                   Ontario            2011.01 L0A 2F6                   Ontario            2011.02 L2G 7D3                  Ontario             2011.01 L2G 7D3                  Ontario             2011.02 WANT: POSTALCODE        City                   Date          Population M1P 3F7                 Toronto             2011.01         1,931,796   <- 5,795,389 divided by the 3 different Postal Codes in Toronto within the same date M1P 3F7                 Toronto             2011.02         2,901,532   <- 5,803,064 divided by the 2 different Postal Codes in Toronto within the same date       M1J 2G8                  Toronto            2011.01         1,931,796   <- 5,795,389 divided by the 3 different Postal Codes in Toronto within the same date M1J 2G8                  Toronto            2011.02         2,901,532   <- 5,803,064 divided by the 2 different Postal Codes in Toronto within the same date M1J 4K9                  Toronto            2011.01         1,931,796   <- 5,795,389 divided by the 3 different Postal Codes in Toronto within the same date L0A 2F6                   Ontario            2011.01         1,339,231   <- 2,678,462 divided by the 2 different Postal Codes in Ontario within the same date L0A 2F6                   Ontario            2011.02         1,339,377   <- 2,678,754 divided by the 2 different Postal Codes in Ontario within the same date L2G 7D3                  Ontario             2011.01        1,339,231   <- 2,678,462 divided by the 2 different Postal Codes in Ontario within the same date L2G 7D3                  Ontario             2011.02        1,339,377   <- 2,678,754 divided by the 2 different Postal Codes in Ontario within the same date ... View more

Hi,  could someone help please in constructing a way I can count from the back of the date series and see how many fields have non-missing information within the last 3 months, 6-months and 12-months. For example (showing only 3mths): If I have these 4 scenarios:                Oct'11    Nov'11     Dec'11 Scen1       value      value      value Scen2        value        .         value Scen3        value        .           . Scen4          .            .           . Scen5          .         value        . Results:             3-mths Scen1      3  Scen2      2 Scen3      1 Scen4      0 Scen5      1 Thank you ... View more

Hi,  I'm trying to check for collinearity (using VIF), so to do variable selection using proc reg.  I need to check more than one x-var at a time, but I'm not geting any results.. but if I do one at a time it works. WORKS: proc reg data=Data; model Price = Var1 / vif tol; run; quit; endrsubmit; DOESN'T WORK: proc reg data=Data; model Price = Var1 Var2 Var3 / vif tol; run; quit; ... View more