Skip to content

turing-complet/python-ouraring

Repository files navigation

Installation

Easiest way is to get it from PyPI:

pip install oura

Project maintenance

If anyone is interested in taking over maintenance of this project, please contact me at [email protected] or submit an issue. Alternatively, feel free to simply fork this repo or create a new one and publish the package under a new name. In that case, I will add a link here. I still plan on doing what I can to keep this up to date, but don't feel I can commit to maintaining a level of quality and responsiveness that you all deserve.

Note about the v2 API

All sections in the readme should apply to the v2 API that is being rolled out, except that the methods will differ, pandas support is less sophisticated, and I haven't tested the v2 clients with the OAuth2 flow. The auth has been updated to pass tokens in the http header, but the usage of the underlying requests-oauthlib library has not been changed.

Enjoy the latest clients as follows (and see docs). All methods except personal_info take a start_date, end_date, and next_token.

from oura.v2 import OuraClientV2, OuraClientDataFrameV2
v2 = OuraClientDataFrameV2(personal_access_token="MY_PAT")

# methods will be named after the url path (see docs linked above)
v2.heartrate()

# pandas methods end with _df
v2.tags_df()

Getting started

Both personal access tokens and oauth flows are supported by the API (and by this library). For personal use, the simplest way to start is by getting yourself a PAT and supplying it to a client:

client = OuraClient(personal_access_token="MY_TOKEN")

If you are using oauth, there are a few more steps. First, register an application Then you can use this sample script to authorize access to your own data or some test account data. It will follow the auth code flow and print out the token response. Make sure to add localhost:3030 to the redirect uris for your app (the port can be changed in the script).

./token-request.py <client-id> <client-secret>

Some sample code is located in the samples directory, maybe it will be useful for you. Maybe it will change your life for the better. Maybe it will cause you to rethink using this project at all. Let me know the outcome if you feel like it.

Business time

If you are writing a real application, use the following pattern. Basically, the work is done by the underlying oauthlib to use the refresh token whenever the access token has expired, and you supply the refresh callback to save the new tokens for next time. This seems to have worked fine for me, but I don't actually use this library that much

from oura import OuraClient, OuraOAuth2Client

auth_client = OuraOAuth2Client(client_id='my_application', client_secret='random-string')
url = auth_client.authorize_endpoint(scope='defaults to all scopes', 'https://localhost/myendpoint')
# user clicks url, auth happens, then redirect to given url

Now we handle the redirect by exchanging an auth code for a token

# save this somewhere, see below
token_dict = auth_client.fetch_access_token(code='auth_code_from_query_string')

Now that's out of the way, you can call the api:

# supply all the params for auto refresh
oura = OuraClient(<client_id>, <client_secret> <access_token>, <refresh_token>, <refresh_callback>)

# or just these for make calls until token expires
oura = OuraClient(<client_id>, <access_token>)

# make authenticated API calls
oura.user_info()
oura.sleep_summary(start='2018-12-05', end='2018-12-10')
oura.activity_summary(start='2018-12-25')

The refresh_callback is a fuction that takes a token dict and saves it somewhere. It will look like:

{'token_type': 'bearer', 'refresh_token': <refresh>, 'access_token': <token>, 'expires_in': 86400, 'expires_at': 1546485086.3277025}

Working with pandas

You can also make requests and have the data converted to pandas dataframes by using the pandas client. Some customization is available but subject to future improvement.

client = OuraClientDataFrame(...)
bedtime = client.bedtime_df(start, end, convert=True)

In [3]: client.bedtime_df()
Out[3]:
              bedtime_window                   status
  date
  2020-03-17  {'start': -3600, 'end': 0} IDEAL_BEDTIME_AVAILABLE
  2020-03-18  {'start': None, 'end': None} LOW_SLEEP_SCORES

Live your life.