newsdev / ap-deja-vu Goto Github PK

It would be great if it was easier to see what's under the hood of a deja vu data folder.

• No duplicate data files
• No initialization data (or isolated as such, like your error response buttons)
• Start all data sets at 0% reporting and end with 100% reporting. I know AP data is messier than this, but this can be our opportunity to make better data for ourselves.
• Ability to upload and share customized data for different test scenarios (zeros, ties, multiple winners, etc.)

Need to have a local and a national folder containing the files.

The utils library wasn't in the requirements.txt file ... easy nuff fix. 😉

(venv-ap-deja-vu) ➜  ap-deja-vu git:(master) python -m ap_deja_vu.app
Traceback (most recent call last):
  File "/usr/local/Cellar/python/3.7.2/Frameworks/Python.framework/Versions/3.7/lib/python3.7/runpy.py", line 193, in _run_module_as_main
    "__main__", mod_spec)
  File "/usr/local/Cellar/python/3.7.2/Frameworks/Python.framework/Versions/3.7/lib/python3.7/runpy.py", line 85, in _run_code
    exec(code, run_globals)
  File "/Users/agorton/Projects/ap-deja-vu/ap_deja_vu/app.py", line 13, in <module>
    import utils
ModuleNotFoundError: No module named 'utils'

It would be nice to avoid keeping initialization data in the deja vu cycle, and instead just add a button to test what happens with initialization data like we do with error codes. It's difficult to know when and were the init data exists in a deja vu collection.

Right now, elex-loader's update.sh script will always push deja-vu beyond position: 0, so it's not actually possible to queue up the position: 0 results.

Sets to zero:
http://campfin.newsdev.net/ap-deja-vu/2016-03-01?position=0&playback=1

Run ./update.sh:
Pushes to position: 1.

This will help with our models, according to @joshkatz.

This one.

For example, getting a fake-data server set up currently requires virtualenvwrapper's add2virtualenv to be available; otherwise one could just run python setup.py develop or pip install ..

This could also help if we ever wanted to release this on PyPI.

It would be great to have a button to load production data on staging.