Test of non-linearity

A bootstrap test of non-linearity for time series data

Description

There are three M-files:
1. third.m, calculates the third-order moment for a time series
2. aaft.m, calculates the Amplitude Adjusted Fourier Transform (for more details see: D. Kugiumtzis, Surrogate data test for nonlinearity including monotonic transformations, Phys Rev E, vol. 62, 1, 2000)
3. boot.m, uses the above two m-files to calculate a bootstrap test of non-linearity. The graphical output from the test is useful for: a) identifying the type of non-linearity, and b)informally checking non-stationarity.

Full details of the test are in the paper: Barnett & Wolff, A Time-Domain Test for Some Types of Nonlinearity, IEEE Transactions on Signal Processing (2005, vol 53, issue 1).

• MATLAB release: MATLAB 5.3 (R11)

[Dmitry Volobuev]

Thank You, it saves my time.

[Asim]

Dear Adrain,
Thank you for sharing the code. I ran tha code with a dataset. However, I could read the out comes properly. is there any clear way of explaining the outcomes. ow to interpret the results.
thanks

[Adrian Barnett]

Dear Asim

Sorry for the delay in replying. Have you tried reading the paper in the 'IEEE Transactions on Signal Processing', vol 53, issue 1, pages 26-33?

There are two parts to the output: i) A graphical part that shows the co-ordinates of the third order moment that are outside the test limits, ii) An overall boostrap test of non-linearity based on the third order moment.

Adrian

[Ali Ali]

Thanks alot. It is a very useful tool

[Min Gan]

 

[Min Gan]

Dear Adrain,
Thank you for sharing the code.
I notice that the codes are bootstraps using AAFT mehtods.
Could you please send me this nonlinearity test method code with CAAFT?
thanks very much!
Gan