InfoCrit

This is a simple function that I ended up using quite often.

InfoCrit() returns a vector containing the AICs or BICs for autoregressive models of a given variable for lags 0:n .

Simple usage instructions

1.Download a copy of this function onto your machine

2.Pull the function into your R environment with source("your\path\InfoCrit")

3.The function can now be called like any other InfoCrit(data_, column, lags, Type)

Parameters

• data_

Expects a dataframe containing atleast 1 named variable

• column

String - Name of the column of the dataframe to be used in the AR model.

• lags

int - Number of lags in the highest order model you want to run.

• Type

String - Default = "AIC" , but also allows "BIC".At some point I may add others.

Technical Considerations

Formula used to calculate the AIC:

Formula used to calculate the BIC:

Where

• 

• T is the total number of time periods

This may not agree with other R functions used to calculate Information criterion as they sometimes use different formulas.

Acknowledgments

Wesely Bundell Ph.D. http://www.wesleyblundell.com/ for assigning the homework that made this necessary.