Get Started
===========

Initialization
--------------

The basic usage of `idinn` starts with a sourcing model and a controller. First, initialize a sourcing model, such as :class:`idinn.sourcing_model.SingleSourcingModel`, with your preferred parameters.

.. doctest:: python

    >>> from idinn.sourcing_model import SingleSourcingModel
    >>> from idinn.demand import UniformDemand

    >>> # Initialize the sourcing model
    >>> single_sourcing_model = SingleSourcingModel(
    ...     lead_time=0,
    ...     holding_cost=5,
    ...     shortage_cost=495,
    ...     batch_size=32,
    ...     init_inventory=10,
    ...     demand_generator=UniformDemand(low=1, high=4),
    ... )


Afterwards, initialize a controller that is compatible with the chosen sourcing model. In the above single-sourcing example, the relevant controller is :class:`idinn.single_controller.BaseStockController`. It is also possible to use other controllers that solve single-sourcing problems, such as :class:`idinn.single_controller.SingleSourcingNeuralController`.

.. doctest:: python

    >>> from idinn.single_controller import BaseStockController
    >>> # Initialize the controller
    >>> controller = BaseStockController()

Training
--------

The selected controller needs to be trained to find suitable parameters.

.. doctest:: python

    >>> # Train the controller
    >>> controller.fit(sourcing_model=single_sourcing_model)

Plotting and Order Calculation
------------------------------------------

After completing training, we can inspect how the controller performs in the specified sourcing environment by plotting the inventory and order evolution.

.. doctest:: python

    >>> # Simulate and plot the results
    >>> controller.plot(sourcing_model=single_sourcing_model, sourcing_periods=100)  # doctest: +SKIP

.. image:: ../_static/single_sourcing_output.svg
   :alt: Output of the single sourcing model and controller
   :align: center

The trained controller can be used to predict order quantities.

.. doctest:: python

    >>> # Predict order quantity for a given system state
    >>> controller.predict(current_inventory=10, past_orders=[1, 5])
    0