search

Acicone / MvFIF

Multivariate Fast Iterative Filtering
MIT License
7 stars 7 forks source link

I am confused about the parameter setting of the function 'MvFIF_v7'. #1

Open zhangyi84688 opened 3 years ago

zhangyi84688 commented 3 years ago

I set the parameters of the function 'MvFIF_v7' to perform multi-channel signal decomposition, but the results obtained are very different from the actual results. In your paper entitled 'Multivariate Fast Iterative Filtering for the decomposition of nonstationary signals', how do you set the parameters of the function 'MvFIF_v7' in the robustness test of MvFIF against noise and the analysis of the filter bank characteristics of Gaussian white noise? I look forward to your sharing an example of 'MvFIF_v7' for filter bank analysis of Gaussian white noise or for noise robustness analysis. Looking forward to your reply!

Best regards

Acicone commented 3 years ago

Dear Yi Zhang, thanks for your interest in the code. I assume you are refering to Figure 8 and Figure 5 from 'Multivariate Fast Iterative Filtering for the decomposition of nonstationary signals', right? For the white Gaussian noise I use the folliwing settings: opts=Settings_FIF_v3('verbose',0,'NIMFs',8,'alpha','ave','Xi',1.70*2,'delta',0.001); For the robustness test, instead, opts=Settings_FIF_v3('verbose',0,'NIMFs',7,'alpha',100,'Xi',1.8); In this second case you want to extend first the signals using the Extend_sig_v2 function from the FIF folder. Feel free to contact me if you have any other questions. Best,

Antonio

zhangyi84688 commented 3 years ago

Dear Antonio, Thank you for your reply. I performed a test with a 3-channel signal based on this parameter you provided. In the test, when I don't add noise, the 3 components obtained are more realistic, and when noise is added, its decomposition becomes quite bad. And the decomposition result is quite related to the component set, I don't know how to get the number of components of decomposition adaptively. Also, in your paper titled 'Multivariate Fast Iterative Filtering for the decomposition of nonstationary signals', I would appreciate it if you could share the MATLAB m-file about the filter bank characteristics of Gaussian white noise and the noise robustness test. Looking forward to your reply!

Best regards

Acicone commented 3 years ago

Dear Yi Zhang, I plan to publish the code on my github as soon as the associated paper gets accepted in the journal I submitted it. Best regards

zhangyi84688 commented 3 years ago

Dear Antonio, Thank you for your reply. I sincerely hope your paper will be accepted soon. Best regards