Added ability to report results to a URL, and sample server code for saving it to a database

[rishsriv]

No description provided.

[wongjingping]

Made a few key changes:

• Separated cloud function for writing to Bigquery / Postgres into different folders. Users can distinguish which cloud function they want to deploy more easily with worrying about bringing in excessive deps. This also facilitates easy switching between reporting endpoints if necessary.
• Pass all args into upload_results, and let the endpoint decide which ones it wants to keep. By passing a superset we allow users to pick and choose which metrics are meaningful to them.
• Added instructions for deploying and testing the cloud function. A little bit more work is required for Bigquery due to the service account credentials required (see pandas_gbq docs).
• Fix df.to_gbq(...) calls to pandas_gbq.to_gbq(df, ...), which is the updated way to perform table writes (see doc).

Tested it by deploying both BQ and Postgres versions and have verified that it works.

[wongjingping]

Noice thank you for these improvements!