Cleanup the notebooks and make correspondence to papers figures clear

[JesseLivezey]

• Fig 1
• Fig 2
• Fig 3
• Fig 4

[JunwenBai]

Hi, Jesse,
Have you cleaned up the notebooks ever? I think most of the ipython notebooks *.ipynb fail to run. Do you mind looking into them? Thx
Best

[JesseLivezey]

Thanks for bumping this. I'll try and make progress in the next few weeks. I should be able to get at least the Fig 1 notebook cleaned up fairly quickly.

[JunwenBai]

Thank you so much! I tried to reproduce the Fig 1 results using lorenz_fig.ipynb, but DCA was giving almost the same results as PCA. T_T Many thanks for working on this!
Btw, do you mind also providing the datasets or the links to the datasets for other notebooks? e.g. sabes, neural data. Most notebooks are not working basically due to 3 reasons: 1. CCA instead of DCA. And simply replacing CCA with DCA is not solving the problem. 2. Lack of datasets. Many datasets still have special data paths. 3. Deprecated functions. For instance, function random_basis has a new argument rng, while many ipynbs still overlook this arg.
Anyway, really appreciate it if you can clean up the codes. The paper is quite intriguing!