timd a simplified way to time your python applications

Timd provides a simple way of timing segments of your code by providing a single class "Measure" that does two things: 1. It writes the execution duration of your code as a log message (at info level) 2. It stores the execution time for later evaluation

Installation

You can install timd using pip:

pip install timd

Examples

For the examples below, I'll pipe stderr to stdout, so that the below examples can be kept up to date using python doctest. If you want to try out timd for yourself, you don't have to do this.

>>> import sys
>>> sys.stderr = sys.stdout

Prepare for the examples: Set logging level to INFO

For our examples, we want the logging module imported and we want logging set to info level

>>> import logging
>>> logging.basicConfig(level=logging.INFO)

Example 1: Using context manager for logging

timd is best illustrated using examples

>>> from timd import Measure
>>> with Measure():
...     x = sum([4]*1000)                               # doctest: +ELLIPSIS
INFO:root:Execution took ...s

Example 2: Using decorator version for logging

We can also decorate functions with a Measure:

>>> from timd import Measure
>>> @Measure()
... def myfunction():
...     return sum([4]*1000)
>>> print(myfunction())                                 # doctest: +ELLIPSIS
INFO:root:Execution took ...s
4000

Note that every time my function is called, we will receive a log message.

Example 3: Using multiple `Measure` objects to track times

In some cases, we want to track times associated with different types of operations. This following example illustrates a more complex use of different Measure objects:

>>> from timd import Measure
>>> logger = logging.getLogger('times')
>>> trackers = {'printing': Measure('Printing took {}s', logger),
...             'logical operations': Measure('Logical operation took {}s',
...                                           logger)}
>>> with trackers['printing']:
...     print('Starting')                               # doctest: +ELLIPSIS
Starting
INFO:times:Printing took ...s
>>> t = trackers['printing']
>>> @t
... def show_formatted(c):
...     print('The condition was {}'.format(c))
>>> for x in range(3):
...     with trackers['logical operations']:
...         condition = x % 3 == 0
...     show_formatted(condition)                       # doctest: +ELLIPSIS
INFO:times:Logical operation took ...s
The condition was True
INFO:times:Printing took ...s
INFO:times:Logical operation took ...s
The condition was False
INFO:times:Printing took ...s
INFO:times:Logical operation took ...s
The condition was False
INFO:times:Printing took ...s

Note that we defined a separate logger for our time measurements, so that we can use the functionality of the logging module to send timing statements to e.g. a separate logfile. Furthermore, each logger stores information about it tracked:

>>> total_printing = sum(trackers['printing'].times)
>>> total_logic = sum(trackers['logical operations'].times)
>>> print('Time spent printing: {}s, Time spent with logic: {}s'.format(
...     total_printing, total_logic))                   # doctest: +ELLIPSIS
Time spent printing: ...s, Time spent with logic: ...s

Tests

Test status on master:

igordertigor/python-timd