.-
help for ^superxt^
.-

Random-effects estimation commands: xt
--------------------------------------

	^superxt^ command depvar varlist [^if^ exp] [^in^ range] ^,^ 
			[ ^i(^varname^)^ ^q^uadrat^(^#^)^ ^l^evel^(^#^)^ ^t^ype^(^#^)^ ^r^ho^(^#^)^  ]

This command shares the features of all estimation commands; see help @est@.

To reset problem-size limits, see help @matsize@.


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

^superxt^ estimates commands from the xt family of commands which uses random-effects. For example: xtprobit, xtlogit, xttobit... It starts by the default command and then attempts other starting values and number of quadrature points in case  of a failure to converge.


Options
-------

^i(^varname^)^ specifies the variable corresponding to an independent unit
    (e.g., a subject id).  ^i(^varname^)^ is not optional.

^quadrat(^#^)^ specifies the number of points to use for Gaussian-Hermite
    quadrature. It is optional, and the default is 12. Increasing this value
    improves accuracy, but also increases computation time.  Computation time
    is roughly proportional to its value. Changing ^quadrat(^#^)^ only affects
    results if ^option(^#^)^ is set to 2.

^level(^#^)^ specifies the confidence level, in percent, for confidence
    intervals.  The default is ^level(95)^ or as set by ^set level^.

^type(^#^)^ is set to 1 when you want superxt try possibilities, starting
    with the default, until it finds a solution or runs out of options. When
    ^type(^#^)^ is set to 2, then superxt estimates the command with a starting
    value from the equivalent non-xt command and rho.

^rho(^#^)^ sets the rho you want to use as a starting value, the default is 0.5.

Examples
--------

   . ^superxt probit y x, i(id)^
   . ^superxt logit y x, i(id) rho(0.3)^
   . ^superxt tobit y x, i(id) quad(29) type(2) ul(0) ll(100)^
   . ^superxt^


Author
------

	Guillaume R. Frechette
	http://www.people.hbs.edu/gfrechette/html/econ.htm


Also see
--------

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