.-
help for ^arimafit^                           
.-

Criteria for assessing ARIMA model fit --------------------------------------

^arimafit^

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

^arimafit^ estimates Akaike's Information Criterion -AIC- (Akaike 1974) and Schwarz's Criterion -SIC- (Schwarz 1978) for assessing maximum likelihood estimates model goodness of fit.

This command can only be run after the @arima@ command, applied to a single dependent variable.

This code is based on ^mlfit^ by Aurelio Tobias and Michael J Campbell, published in STB-45 (sg90).

The command defines the scalars ^np^ (number of estimated parameters), ^llf^ (minus twice the log of the likelihood), ^aic^ and ^sic^ for later use.

Examples --------

. ^arima m1, arima(1,1,1)^ . ^arimafit^

. ^arima m1, arima(1,1,1)^ . ^qui arimafit^ . ^dis r(llf)^ . ^dis r(aic)^

Reference ---------

Akaike, H. (1974). A new look at statistical model identification. IEEE Transactions on Automatic Control, 19: 716-722.

Schwarz, G. (1978). Estimating the dimension of a model. Annals of Statistics, 6: 461-464.

Authors ------- Christopher F Baum Boston College, Chestnut Hill MA 02467 USA baum@@bc.edu

Also see --------

On-line: ^help^ for @arima@