.-
help for ^intterms^ (statalist distribution 27apr1998)
.-
Generate all indicators for all intersections of two categorical variables
- --------------------------------------------------------------------------
^intterms^ variable1 variable2 [^if^ exp] [^in^ range] , ^gen(^var_stub^)^
[^, nodrop^ ^full^ ^index^ ]
Description
- -----------
^intterms^ generates a set of indicators for all intersections of two
categorical variables.
Required "option"
- -----------------
^gen(^var_stub^)^ designates the "variable stub" which will be used as the
leading characters in the indicator variable names. Note, use only a
single character if you wish to allow the maximum number of digits for use
in representing the categories -- 6 digits.
Options
- -------
^nodrop^ indicates that the first indicator variable *NOT* be dropped from
the set of generated variables. This option creates a set of indicators
that fully span the space and are collectively collinear.
^full^ indicates that indicators should be generated for all combinations of
the two variables, even if the indicator is constant for a combination.
^index^ specifies that the "index" values of the categories be used in naming
the indicator variable rather than the values of the variable. For
example, if ^cat1^ takes on the values 1, 2, 5, 6 and "D" were specified
on the ^gen()^ option; the indicator variables would normally start
with D1... , D2..., D3...., D5...., D6.... . With the ^index^ option,
the indicators would be named D1... , D2..., D3...., D4...., D5.... .
Examples
- --------
. intterms mpg foreign, gen(D)
. intterms mpg foreign, gen(ivar) index
Author
- ------
Vince Wiggins
StataCorp.
vwiggins@@stata.com
Also see
- --------
Manual: ^[U] 26 Estimation and post-estimation commands^
^[U] 35 Overview of model estimation^
^[R] xi^
On-line: ^[R] xi^, help for any estimation command