SAAT-API

This repository contains code for a Flask API server for persisting and fetching heartbeat and subjective psychological data from a Postgres database via the Psycopg2 and Pandas Python packages.

Running the server on a local VM w/ Vagrant

You'll need Vagrant and as well as a virtual machine provider such as VirtualBox installed. In the project directory (where the vagrant file is), run vagrant up to boot up the machine. The first time you run this will take several minutes as the base machine image is downloaded and the necessary components are provisioned (Postgres is installed, bleeding edge Python is downloaded and built).

Once the above is complete, you can issue vagrant ssh to shell into the running virtual machine. The /vagrant directory on the VM is synced by Vagrant to mirror the vagrant file directory on your machine. To start the server, issue:

cd /vagrant
python app.py

The debug server will automatically reload changed code, which you just edit on the synced directory on your host machine with your editor of choice. There is a Postgresql instance on the virtual machine that the server will interact with by default if you don't have an instance/config.py to override. Note that you may need to alter the port forwarding on the Vagrantfile to allow interaction with remote databases.

Visit http://localhost:5000/ in a web browser to access the server or use curl as shown below.

To run tests (these tests will destroy the data on the VM database!) issue python app_tests.py from the /vagrant directory.

To shutdown the VM, issue vagrant halt

To delete the VM entirely (except for the downloaded base image), issue vagrant destroy

Running the server locally

You'll first need Postgres 9.6 installed on your machine. If you are on OSX and use Homebrew, just do brew update then brew install postgres.

Next create a postgres database and load the schema:

# Connect to management database to enter the psql prompt
psql postgres
# Create the database
create database saatdb01;
# Disconnect to leave psql
\q
# Load the schema
psql -d "postgresql://@localhost:5432/saatdb01" < schema.sql

Then you'll want to install virtualenv to manage Python dependencies. Create a Python 3.6 virtual environment named venv and then use the new environment's pip to install dependencies like so:

venv/bin/pip install -r requirements.txt

With Flask installed, the following commands can be used to start up a reloading server in debug mode.

export FLASK_APP=app.py
export FLASK_DEBUG=true
export PG_USER=$(whoami)
flask run

You will need to create a local instances folder with a configuration like so:

mkdir instances
cp config.py instances

Edit the values in config.py to reflect your local database credentials.

Interacting with the server

The curl commands below demonstrate basic interaction with the API. RR interval data can be created or fetched with POST and GET requests respectively. See the below referenced JSON files for examples.

curl -i -X POST localhost:5000/users/watson/measurements/rr_intervals -H "Content-Type: application/json" --data-binary "@sample_post_single_rri.json"

curl -i -X GET localhost:5000/users/watson/measurements/rr_intervals -H "Content-Type: application/json" --data-binary "@sample_get_rri.json"

Troubleshooting Dependencies

If you're getting cryptic "Reason: image not found" errors when trying to import psycopg2, you might need the following fix.

export DYLD_FALLBACK_LIBRARY_PATH=$HOME/anaconda3/lib/:$DYLD_FALLBACK_LIBRARY_PATH

See this stack overflow question for details.

Connecting to the prod DB

The app loads postgres connection params from environment variables. If you'd like to connect to the prod database, find the credentials in the file saat-api.sh in the SAAT-Internal repo.

Pre load them into your shell:

source saat-api.sh

then run the app:

python app.py

You can also run it like this:

PG_HOST=bla.bla.us-west-2.rds.amazonaws.com PG_DB=YourSweetDB PG_USER=YourCoolUser PG_PASS=Blabla python app.py