Sabre manual

Syntax: DATA list of variables

The DATA command has two purposes. Its primary purpose is to introduce a list of variables which are to be read by a subsequent READ command. The length of the variables is determined dynamically from the data introduced by the READ command. All variables are read as covariates; they can be subsequently converted to factors by the FACTOR command or transformed by the TRANSFORM command.

The variables to be read will normally include the following:

  1. A case variable. This gives the case number of the observation. By default, this is taken to be the first variable given in the list of variables, but this may be over-ridden by using the CASE command.

  2. A y-variable, taking the values 0 and 1 only for a binary model, or non-negative integer values for a Poisson model.

  3. A set of covariates and factors.

If the command is used more than once in a SABRE session, the second and subsequent uses will introduce a new dataset to be analysed, and will also clear the workspace, delete all model fitting information and reset the model constants to their default values, preparing the program for an entirely new analysis.

Sabre manual

Go to: Sabre home page | Sabre manual | Downloading & Installing Sabre | Sabre examples | Training materials | Sabre mailing list | Contact us

Other links: Centre for e-Science | Centre for Applied Statistics