threedi-connector

Python client library for interacting with the 3Di API.

Install

$ pip install git+https://github.com/nens/threedi-connector.git

Use

Import the threedi_connector module:

>>> import threedi_connector

Start a simulation:

>>> simulation = threedi_connector.start_simulation(
...    model_slug='v2_bergermeer-v2_bergermeer_bres_maalstop-55-784c561ecebf9433cd7beb8b6a22a14f2841cda4',
...    organisation_uuid='61f5a464c35044c19bc7d4b42d7f58cb',
...    scenario_name='test-lib (this result can be deleted)',
...    start='2016-10-18T00:00',
...    end='2016-10-18T00:30')
Username: <your username>
Password: <your pw>

Get simulation task metadata using the returned Simulation object:

>>> simulation.info
{u'organisation_id': u'61f5a464c35044c19bc7d4b42d7f58cb',
 u'result_id': u'300f9ed8d1d54ff8b4def220a81060c8',
 u'status': u'received',
 u'subgrid_id': u'headless:1e0a2077d55044399a857c766e328645'}

Information from 3Di API can be requested using the Client. For example, getting the saved states and queued simulation:

>>> client = threedi_connector.Client()
>>> client.authenticate()  # authenticate once for this session
Username: <your username>
Password: <your pw>
>>> client.saved_states
[{'some': 'data', ...}, ...]
>>> sm.queued_tasks
[{'more': 'data', ...}, ...]

Interactive mode

Interactive mode is on by default. You can disable interactive prompts by authenticating using the .authenticate(username, password) method on Client or Simulation objects. Note: if you call .authenticate() without arguments, it will prompt for credentials.

To disable interactivity completely it is safer to just instantiate with interactive=False:

>>> client = threedi_connector.Client(interactive=False)
>>> simulation = threedi_connector.Simulation(interactive=False)

Running tests

$ python setup.py test

Design philosophy

Be good at doing one (or maybe two or three) thing(s).

A very simple API, with clear public and private classes/methods/functions. A simple API means also less need for thorough documentation.

As few dependencies as possible for maximum portability (ideally: 0; the current situation).

Designed for both Jupyter notebook use (therefore i.e. interactive authentication), and use as a library.