Don't download data from ourselves

[duboism]

Tutorials 2.3 and 2.3 download data... from this repository. This is useless since the data are checkout out (in fact, those data represent roughly 50% of the size of the repository).

I see 2 options (by ascending level of complexity):

1. Keep the data in the repository and just open them
2. Don't put the data in the repository but store them somewhere (and download them) eventually through Git LFS
3. Include the data in the archive but not in the repository (this way, no need to download them at runtime)

Solution 2 and 3 are more satisfying but harder to do. I feel solution 1 is enough.

[martinberoiz]

I agree that for users working on local laptops that have the repo checked out, downloading them again is silly, but I'm worried about how would it work on Colab or mybinder.

I tried opening the file in Colab without downloading it first (I don't know where it's stored) and it didn't work for me.
I tried doing file_name = "PyCBC_T2_1.gwf" and file_name = "./Data/PyCBC_T2_1.gwf", both didn't work.

If we find a way to make it work online I think it's a good idea.

[martinberoiz]

A possible solution may be adding:

if os.path.exists(f"./Data/{file_name}"):
    file_name = "./Data/.."
else:
    # download the file
    ...
    file_name = ...

The above is just pseudo code, hopefully it's clear enough.

Also, maybe we should be careful with OS paths because we may have students working on Windows (i.e. use os.path.join everywhere)

[duboism]

You're absolutely right, I forgot about Colab :) Silly me!

The solution you propose is fine but I would just download the file in the Data directory (creating it if needed). I might open a PR on this on Monday.

[duboism]

See #49.