Iteround: Sum-safe Rounding for Iterables

.. image:: https://travis-ci.org/cgdeboer/iteround.svg?branch=master :target: https://travis-ci.org/cgdeboer/iteround

.. image:: https://img.shields.io/pypi/v/iteround.svg :target: https://pypi.org/project/iteround/

.. image:: https://img.shields.io/conda/vn/conda-forge/iteround.svg :target: https://anaconda.org/conda-forge/iteround

Iteround is an organic (standard library) sum-safe rounding library for Python iterables (lists, tuples, dicts).

.. image:: https://raw.githubusercontent.com/cgdeboer/iteround/master/docs/iteround.png

Example Code:

.. code-block:: python

>>> import iteround
>>> data = {'foo': 60.19012964572332,
            'bar': 15.428802458406679,
            'baz': 24.381067895870007}
>>> sum(data.values())
100.0
>>> rounded = iteround.saferound(data, 0)
>>> rounded
{'foo': 60.0,
 'bar': 16.0,
 'baz': 24.0}
>>> sum(rounded.values())
100.0

How It Works

Iteround provides a single method, called :code:saferound that takes the following inputs:

iterable (list, dict, set, numpy.array, generator): list(y) of numbers If a dict is passed in, the values must be all floats.

places (int): Places for rounding. Number of places each item in the set should be rounded to.

topline (float, optional): Topline to match Useful in places where we want the total sum to match a different topline than the sum of iterable. This can useful in cases where original values are altered before passing into the saferound method, but the original sum needs to be maintained.

strategy (str, optional): The strategy used to clean up rounding errors 'difference', 'largest', 'smallest'. Defaults to 'difference'

'difference' seeks to minimize the sum of the array of the
differences between the original value and the rounded value of
each item in the iterable. It will adjust the items with the
largest difference to preserve the sum. This is the default.

'largest' for any post rounding adjustments, sort the array by
the largest values to smallest and adjust those first.

'smallest' for any post rounding adjustments, sort the array by
the smallest values to largest, adjust the smaller ones first.

Strategy strings are available as:
    :code:`iteround.DIFFERENCE`
    :code:`iteround.LARGEST`
    :code:`iteround.SMALLEST`

If 'dict' or 'tuple' are passed, result will be dict or tuple. All other iterables (range, map, np.array, etc) will return list.

Feature Support

Iteround definitely supports at least these iterables.

• list
• tuple
• dict
• OrderedDict

Iteround officially supports Python 2.7 & 3.4–3.6.

Installation

To install Iteround, use pipenv <http://pipenv.org/>_ (or pip, of course):

.. code-block:: bash

$ pipenv install iteround

Documentation

Documentation beyond this readme will be available soon.

How to Contribute

#. Check for open issues or open a fresh issue to start a discussion around a feature idea or a bug. #. Fork the repository_ on GitHub to start making your changes to the master branch (or branch off of it). #. Write a test which shows that the bug was fixed or that the feature works as expected. #. Send a pull request. Make sure to add yourself to AUTHORS_.

.. _the repository: https://github.com/cgdeboer/iteround .. _AUTHORS: https://github.com/cgdeboer/iteround/blob/master/AUTHORS.rst