.. _sec_simple_example:

Simple example
==============

To illustrate the PopXL concepts, here is a simple example.

You need to import the ``popxl`` package in order to use it:

.. code-block:: python

  import popxl


The IR in PopXL is represented by the class :py:class:`~popxl.Ir`.
:numref:`simple_example` contains a basic example of how to construct such an object.

.. literalinclude:: files/simple_addition.py
  :language: python
  :lines: 8-27
  :name: simple_example
  :caption: Simple example using PopXL
  :linenos:
  :lineno-match:

.. only:: html

    :download:`Download simple_addition.py <files/simple_addition.py>`

In PopXL, an IR is essentially a collection of :py:class:`Graph <popxl.Graph>` objects.
Each such graph contains a number of operations.
Each IR has a *main graph* that is constructed by default.
This main graph serves as the entry point for your model and is created with :py:func:`~popxl.Ir.main_graph` in :numref:`simple_example`.

By adding operations within a ``with main`` context, the operations
are automatically added to the main graph.
In this example, three operations are added: ``host_load``, ``add`` and ``host_store``.

In :numref:`simple_example`, we created two device-to-host streams ``input0_stream`` and ``input1_stream``
and one host-to-device stream ``output_stream``.
The ``host_load`` operations are used to stream data from the host to
the device populating tensors ``a`` and ``b``, respectively.
Another operation, ``add``, then adds these two tensors together.
Finally, the ``host_store`` streams the result data back from the device to the host.