PROC PANEL writes the transformed series to an output data set. That is, if the user selects FIXONE, FIXONETIME, or RANONE and supplies the OUTTRANS = option, the transformed dependent variable and independent variables are written out to a SAS data set; other variables in the input data set are copied unchanged.

Say that your data set contains variables y, x1, x2, x3, and z2. The following statements result in a SAS data set:

   proc panel data=datain outtrans=dataout;
      id cs ts;
      model y = x1 x2 x3 / fixone;
   run;

First, z2 is copied over. Then _Int, x1, x2, y, and x3, are replaced with their mean deviates (from cross sections). Furthermore, two new variables are created.

_MODELL_

is the model’s label (if it exists).

_METHOD_

is the model’s transformation type. In the FIXONE case, this is _FIXONE_ or _FIXONETIME_. If the model RANONE model is selected, the _METHOD_ variable is either _Ran1FB_, _Ran1WK_, _Ran1WH_, or _Ran1NL_, depending on the variance component estimators chosen.