=========================================================
Tutorial of *OperationCount*
=========================================================

| *OperationCount* class has the aggregated log as the member variable and has the method functions to get the information.

.. image:: operation_count.png
 :scale: 70%
 :align: center

.. code-block:: python

    import openLA as la

    course_info = la.CourseInformation(files_dir="dataset_sample", course_id="A")
    event_stream = course_info.load_eventstream()

    operation_count = la.convert_into_operation_count(event_stream,
                                                      user_id=None,        # all user
                                                      contents_id=None,    # all contents
                                                      operation_name=None  # all operation
                                                      )


| The method functions in below table are available now.
| You can get the information of the log like ``operation_count.num_users()``.
| The documentation is in :doc:`../scripts/operation_count`

+-----------------------------+------------------------------------------------------------+
| function                    | description                                                |
+=============================+============================================================+
| num_users                   | Get the number of users in the log                         |
+-----------------------------+------------------------------------------------------------+
| user_id                     | Get the unique user ids in the log                         |
+-----------------------------+------------------------------------------------------------+
| contents_id                 | Get the unique contents ids in the log                     |
+-----------------------------+------------------------------------------------------------+
| operation_name              | Get the unique operation name in the log                   |
+-----------------------------+------------------------------------------------------------+
| operation_count             | Get the count of each (or specified) operation in the log  |
+-----------------------------+------------------------------------------------------------+

If you want to process other than the above functions, you can get DataFrame type event stream by ``operation_count.df`` and process with Pandas library.

.. code-block:: python

    import OpenLA as la
    import pandas as pd

    operations = ["NEXT", "PREV", "ADD MARKER"]

    course_info, event_stream = la.start_analysis(files_dir="dataset_sample", course_id="A")
    operation_count = la.convert_into_operation_count(event_stream,
                                                      operation_name=operations)

    operation_count_df = operation_count.df
    print(operation_count_df)
    """
        userid contentsid   NEXT   PREV  ADD MARKER
    0       U1         C1  299.0  136.0         2.0
    1       U1         C2  182.0  143.0         0.0
    2       U1         C3  133.0   39.0         0.0
    ...     ...       ...   ...     ...         ...
    """

    # normalize the count of operations
    def normalize(col):
        return (col - min(col)) / (max(col) - min(col))

    operation_count_df.loc[:, operations] = operation_count_df.loc[:, operations].apply(normalize, axis=0)
    print(operation_count_df)
    """
        userid contentsid      NEXT      PREV  ADD MARKER
    0       U1         C1  0.404054  0.303571    0.007435
    1       U1         C2  0.245946  0.319196    0.000000
    2       U1         C3  0.179730  0.087054    0.000000
    ...     ...       ...       ...     ...         ...
    """


Example
=============

.. toctree::
   :maxdepth: 1

   ../jupyter_examples/example_operation_count