Typical Set/Curse of Dimensionality - as Jupyter Notebook


#1

As I was reading through Bob’s excellent case study on typical sets and curse of dimensionality which was in R, I translated the R notebook to jupyter-notebook as I am more familiar with Python libraries. Pls find the notebook at https://github.com/Arvinds-ds/Stan-Code/tree/master/python%20notebooks/curse_dims. It may help any fellow python users to follow along.


#2

That’s fantastic, we love getting outside contributions and translating examples to all the interface language is one area that’s impossible to keep up with.


#3

Looks great.

The seaborn library was not really used. In this case you could do plt.style.use('seaborn-darkgrid').

Available styles can be found with plt.style.available


#4

Thanks. Fixed it


#5

Why the title has Hyperbal even of the plt.title has Hyperball?


#6

Notebook was probably edited inplace (in github).


#7

I missed running the cell. It has been fixed


#8

Thanks! Looks great. Would you mind if I put a copy up on our web site (with attribution, of course). All I need is:

  1. change the author from me to you; you can cite that it’s a derived work from my original and cite that you copied a lot of the text.

  2. include an indication that the Python code is copyright to you and you allow it to be licensed under the BSD for others.

Your own code and your translation of any text is automatically copyright assigned to you, so we can’t use it without your permission in the form of a clear open-source license.

P.S. Of course, I didn’t invent any of this stuff and @betanalpha was really helpful in improving my understanding.


#9

Glad you found it useful and thanks for considering my contribution. I have updated the notebook at https://github.com/Arvinds-ds/Stan-Code/tree/master/python%20notebooks/curse_dims. Also, there is nothing original of mine in here other than translating. Pls feel free to modify it as necessary to meet quality standards.


#10

Thanks for creating this. I posted the case study on our web site with the author “Aravind S” as that’s the most specific name you provided. It does link to your GitHub, but that’s not any more specific about your identity.

Copyright is about quality, not quantity. As soon as you modify the original (for instance, changing “R” to “Python”), you have created a derived work, whose copyright is immediately assigned to you (you don’t need to file anything or even write "copyright on it—it’s automatically assigned to you through the act of creating the derived work).

Now, in order for someone else to use your intellectual property legally, you need to provide a license.

Now, if it’s a derived work, that doesn’t necessarily give you the right to distribute it or license it to others for use—you can’t change a word of a copyrighted novel released under a proprietary license (for instance, through printing a book) and then redistribute the novel as your own work. But in this case, the original was CC-BY, so you can legally redistribute your derived work.

I’m going to assume that your intent was to assign your copyright for the derived work to to me, as that’s what you wrote. If that was not your intent, or if you could do me the favor of adding a surname, please create a pull request for a new version on our web site (under stan-dev organization).


#11

Here’s the link so nobody else has to go link fishing on our rather click-heavy web site:

http://mc-stan.org/users/documentation/case-studies.html


#12

Yes. My intent is to assign copyright for the derived work to you.


#13

Thanks for making that clear.