This obsoletes #3 . data_preparation.py and data_upload.py (just a name) should be separated. The latter is only relevant to us but the former can be used by others who are interested in all intermediate data.
For data upload:
Use the bulk upload mechanism of NDB.
Deploy the combined CSV directly
Host the combined CSV somewhere
Bigger question is whether to stick to this backend, #11 . Say, use Google fusion table to host it directly, like #15 ?
hxu
commented
10 years ago
This obsoletes #3 .
data_preparation.pyanddata_upload.py(just a name) should be separated. The latter is only relevant to us but the former can be used by others who are interested in all intermediate data.For data upload:
Bigger question is whether to stick to this backend, #11 . Say, use Google fusion table to host it directly, like #15 ?