Use API instead of S3 to get the current dataset .zip

[xanderdunn]

Bounty: 10NMR

[gronnbeck]

So, all I'm struggling to find out where the zip files are located. I'm able to use the GraphQL API and fetch relevant data but having a hard time building the HTTP URL to fetch the zip file. Is it part of the task to try to reverse engineer that?

[xanderdunn]

wget --verbose https://api.numer.ai/competitions/current/dataset

[sovaa]

$ wget --verbose https://api.numer.ai/competitions/current/dataset
--2017-12-01 23:03:23--  https://api.numer.ai/competitions/current/dataset
Resolving api.numer.ai (api.numer.ai)... 104.24.28.86, 104.24.29.86, 2400:cb00:2048:1::6818:1c56, ...
Connecting to api.numer.ai (api.numer.ai)|104.24.28.86|:443... connected.
HTTP request sent, awaiting response... 500 Internal Server Error
2017-12-01 23:03:24 ERROR 500: Internal Server Error.

Curl gives the error message:

{"name":"Error","message":"An error occurred. If the problem persists please contact support and try again.","status":500}

What should this issue accomplish? A curl-able url /competitions/current/dataset that returns the zip file? Or a python method in the api lib to get the file?

[sovaa]

API should be on pypi first to be able to do this: https://github.com/numerai/NumerAPI/issues/14

[csp197]

@xanderdunn - Is this bounty still open?
If so, I'd certainly like to resolve it :)