Getting Previous Values of the Same Variable

The LAG and DIF functions can be used within a data step to get the value of a variable from a previous record. The LAG function returns the value of a variable from 'x' records above the current record. When you use the LAG function, a queue is set up so that values of the variable starting at record 'x' are stored. The DIF function works in the same way, except that it returns the difference between the current value and the value 'x' records above it.

The following code demonstrates the use of the two functions:

data t;
  input x;
  first_lag=lag(x);
  second_lag=lag2(x);
  first_dif=dif(x);
  second_dif=dif2(x);
cards;
1
2
4
9
run;

Below is the resulting data set:

     first_    second_    first_    second_
x      lag       lag        dif       dif

1       .         .          .         .
2       1         .          1         .
4       2         1          2         3
9       4         2          5         7

Warning:
Be very careful when using the LAG and DIF functions within conditional sections of a data step, as lag(x) is NOT "the value of X in the last observation"; rather, it is "the value of X last time we executed the statement which calls the lag function".