blaze / blaze

NumPy and Pandas interface to Big Data
blaze.pydata.org
BSD 3-Clause "New" or "Revised" License
3.19k stars 391 forks source link

Support some NoSQL Database #428

Closed mrocklin closed 10 years ago

mrocklin commented 10 years ago

It would be a good experiment to see how well this works. Presumably datashape discovery and such might be more challenging in some cases. Presumably complex Joins and such might not be available. It would be interesting to see what systems are capable of what.

Incidentally I'm interested in the Github data at http://ghtorrent.org . They host a MongoDB view of github's public API data. This would be much easier than going through the json dumps obtained from githubarchive, which aren't the most consistent.

aterrel commented 10 years ago

Yes supporting mongo should be something we do. I'm working with a MongoDB right now with the XDATA demos. I can look into adding it.

BTW, This isn't really an issue, shouldn't discussions like this happen on blaze-dev?

mrocklin commented 10 years ago

http://api.mongodb.org/python/current/api/pymongo/collection.html

mrocklin commented 10 years ago

Working on mongo support now