Could I change a code from jupyternote book to Matlab

Question

Ahmed Gharieb naga 2023-7-7

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1992998-could-i-change-a-code-from-jupyternote-book-to-matlab

回答： Neev 2023-7-11

from fourier_koopman import fourier import numpy as np

x = (np.sin([2*np.pi/24*np.arange(5000)]) + np.sin([2*np.pi/33*np.arange(5000)])).T x = x.astype(np.float32)

f = fourier(num_freqs=2) f.fit(x[:3500], iterations = 1000)

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Joe Vinciguerra 2023-7-7

The short answer is "yes". But you would need to know what the python code does, how it works, and the equivalent matlab syntax. The first couple lines are simple, but that last line might give you trouble if you don't have the documentation for the "fourier_koopman" module, or know how to apply it matlab. Additionally, since it's an iterative fitting algorithm using a deep neural network, you'll likely get a different answer in Matlab. Why don't you just use python, or run the python script from matlab?

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Neev 2023-7-11

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1992998-could-i-change-a-code-from-jupyternote-book-to-matlab#answer_1270968

Hey Ahmed

Yes, the above code you shared can be converted to MATLAB and t=you can find the converted code below:-

% importing the libraries

import fourier_koopman.fourier;

import numpy as np;

t = 0:4999;

x = (sin(2*pi/24*t) + sin(2*pi/33*t))';

x = single(x);

f = fourier_koopman.fourier(num_freqs=2);

f.fit(x(1:3500), 'iterations', 1000);

But, before running the above code, you will have to download the fourier koopman module from the 'Add-on Explorer' in the 'Environment' section at 'Home' of MATLAB taskbar.

You can try the above and let me know oif you need any further help.

I hope the information I shared with you will be of help to you:)