This repository contains the IKATS python api used by core developers and operators contributors to interact with IKATS backend.

• From PyPI: pip install ikats
• From setup.py: python3 setup.py install

You can run the full test campaign with python3 setup.py nosetests A coverage report is also generated and available at cover/index.html

For code quality, IKATS uses pylint. You can run it with pylint --rcfile pylint.rc ikats --ignore="tests"

• ikats/: the api sources
• pylint.rc: configuration file for pylint
• pylint.sh: script that call pylint on specified directory (ikats by default)
• setup.py: install/test/build script
• Tutorial.ipynb: Tutorial as a Jupyter Notebook to know how to play with IKATS

• object: Manipulated set of properties: Timeseries, Dataset, Metadata, Operator, ...
• action: Function operating on an object or a list of objects

There are several ways to interact with API:

• from IkatsAPI
• from an object (located in ikats.objects)

The main structure of the API is the following

o = api.ds.get(name=name)
|   |   |  |   |
|   |   |  |   +----------> Parameters              
|   |   |  +--------------> Actions on the object (get/save/delete/list/...)
|   |   +-----------------> Shortcut of the object to manipulate (ds/ts/md/op/table)
|   +---------------------> IkatsAPI main entrypoint
+-------------------------> output object

Here are some rules dealing with exceptions

• Functional exceptions raised by IKATS api are derived from IkatsException
• All actions may raise exceptions upon errors
• It is possible for the actions returning a status (save, delete, ...) to not raise IkatsException but only a status (useful for simple checks without adding try/except). Just set the parameter raise_exception=False