.-
help for ^reglike^                                  Posted to statalist
14may1998
.-

Obtaining log-likelihood values for linear regression
-----------------------------------------------------

        ^reglike^ [^,^ ^c^ommand^(^commandname^)^ ^df(^#^)^ ]


^reglike^ is run following ^regress^.


Description
-----------

After running ^regress^, ^reglike^ computes the log-likelihood and puts its
value into the global macro ^S_E_ll^.

The ^command()^ and ^df()^ options allow one to trick the ^lrtest^ command into
working when the linear regression is nested within another maximum-likelihood
estimator.


Options
-------

^command(^commandname^)^ specifies the name of another ML estimation command
    with which you wish to use ^lrtest^.  Note that using this option wipes
    out
    the saved results of ^regress^.

^df(^#^)^ specifies the "model" degrees of freedom to be used by ^lrtest^ for the
    linear regression model.  This should be set to whatever number is
    necessary to give the right degrees of freedom for the chi-squared
    statistic from ^lrtest^.


Examples
--------

  . ^regress mpg weight displ^
  . ^reglike^

  . ^xtreg y x1 x2 x3, i(group) mle^
  . ^lrtest, saving(0)^
  . ^reg y x1 x2 x3 if group~=.^
  . ^reglike, c(xtreg) df(2)^
  . ^lrtest^


Author
------

        Bill Sribney
        Stata Corporation
        702 University Drive East
        College Station, TX 77840
        Phone: 409-696-4600
               800-782-8272
        Fax:   409-696-4601
        email: tech_support@@stata.com


Also see
--------

 Manual:  ^[R] lrtest^
On-line:  help for @lrtest@