Time series regression                                        [STB-15: sts4]
----------------------

	^tsfit^ [varlist] [^weight^] [^if^ exp] [^in^ range] [^,^ ^current(^varlist^)^ 

        ^lags(^#[,#,...]^)^ ^nosample^ ^static(^varlist^)^ other-regress-options ]

estimates a time series regression.  Lags of the "varlist" are included in the
regression along with any ^static^ variables.  The current values of variables
in the ^current^ varlist are also included.


Options
-------

^current(^varlist^)^ specifies the RHS variables whose current values are to
  be included.  By default, no current values are included.

^lags(^#[,#,...]^)^ indicates how many lags of each time series to include.  If 
  only one number is indicated, that lag length is applied to the entire 
  varlist.  If the ^lags^ option is omitted, ^lag(0)^ is assumed.  If several 
  numbers separated by commas are typed, the numbers are applied to each of 
  the variables in sequence.  For example, "lags(2,12,4)" would include two 
  lags of the left-hand-side variable, twelve lags of the first explanatory 
  variable, and four lags of the second explanatory variable.  If there are 
  fewer lag lengths than variables, the last lag length listed is used for 
  the remaining variables.  Extra lag lengths are ignored.

^nos^ample suppresses the sample coverage information.  If the ^period^ and
  ^datevars^ have been specified previously, the sample will be displayed 
  before the regression output.

^static(^varlist^)^ specifies the non-time series variables in the regression.


Remarks
-------

^tsfit^ creates any lagged variables needed in the regression and leaves them
in the user's data set.  These lags are needed by commands, such as ^tsmult^,
that can follow ^tsfit^.


Author
------
 
Sean Becketti, Stata Technical Bulletin


Also see
--------

    STB:  sts4 (STB-15)
On-line:  ^help^ for ^datevars^, ^period^, ^tsmult^, and ^tsreg^