Re: ZODB + Jupyter Notebooks

On Tue, Oct 24, 2017 at 10:07 AM, Christopher Lozinski <[email protected]
> wrote:

> There is an obvious marriage to be had between the ZODB and Jupyter
> Notebooks.
>
> For those who do not know Jupyter Notebooks are for data scientists.  You
> get a linear list of
> web pages.


It's more like a sequence of cells, where cells can be computation or text.
Computation cells have output.

...

Right now the JSON is stored as a single file.  No versioning.  No tree of
> related notebooks. Instead people keep multiple related notebooks.  It
> would be reasonably easy to store it in a ZODB.  Then at least you would
> get versions of the JSON file.


Or you could check them into git.

...


> Right now they do not like Github, and do not know how to organize all of
> their different versions of notebooks.
>

<shrug> I don't think any sort of ZODB-based versioning system is going to
fix that.

Note that saved notebooks don't include any data frames generated.  This
has seemed like a bigger problem to me, which is why I created pheather.

https://github.com/jimfulton/pheather

But I'm not really a data scientist, so I'm not the best judge of use cases.

Jim

-- 
Jim Fulton
http://jimfulton.info

-- 
You received this message because you are subscribed to the Google Groups "zodb" group.
To unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
For more options, visit https://groups.google.com/d/optout.