adf inspection steps

There are many methods to test unit roots, such as ADF test, DGFLS test, PP test, KPSS test, ERS test and NP test, but the ADF test method is more commonly used.

ADF test (Augmented Dickey-Fuller test) is a unit root test method in Eviews software to check the stationarity of the series. (Recommended learning: PHP video tutorial)

The output results of the ADF test include the ADF statistics for testing the lagged variable coefficients and the critical values ​​required for the test (1%, 5% , 10%). If the coefficient is significantly different from zero and is actually less than zero, then the hypothesis that it contains a unit root is rejected and the alternative hypothesis of stationarity is accepted.

The steps for doing ADF test are as follows:

1 Test the original time series. At this time, select level for the second item and None for the third item. If it fails the test, explain The original time series is not stationary;

2 Perform the first difference on the original time series and then test it, that is, choose 1st difference for the second item and intercept for the third item. If it still fails the test , you need to perform a quadratic difference transformation;

3 Test of the quadratic difference sequence, that is, select 2nd difference for the second item, and Trend and intercept for the fourth item. Generally, this time series It's stable!

For more PHP related technical articles, please visit the PHP Graphic Tutorial column to learn!

The above is the detailed content of adf inspection steps. For more information, please follow other related articles on the PHP Chinese website!