.. _fc-bench: http://www.math.univ-paris13.fr/~cuvelier/software/Python/fc-bench.html
.. _www.python.org: http://www.python.org/
.. _Miniconda: https://conda.io/miniconda.html
.. _Anaconda: https://www.anaconda.com
.. _Canopy: https://www.enthought.com/product/canopy/
.. _Mayavi: http://docs.enthought.com/mayavi/mayavi/
.. _Matplotlib: https://matplotlib.org/
.. image:: http://www.math.univ-paris13.fr/~cuvelier/software/codes/Python/fc-bench/pyfc-bench_400.png
:width: 400px
:align: center
The **fc\_bench** Python package allows to benchmark functions and much more
Introduction:
-------------
.. _fc\_bench Python package: http://www.math.univ-paris13.fr/~cuvelier/software/Python/fc-bench.html
.. _User's Guide: http://www.math.univ-paris13.fr/~cuvelier/software/codes/Python/fc-bench/0.2.0/fc_bench-0.2.0.pdf
More documentation is available on `fc\_bench Python package`_ dedicated web page with an `User's Guide`_ (pdf).
Installation:
-------------
The **fc\_bench** Python package is available from the Python Package Index, so to install/upgrade simply type
.. code::
pip install fc_bench -U
Testing :
---------
There are demos functions in the ``fc_bench.demos`` module named ``bench_*``.
For example, run the following code under Python:
.. code:: python
from fc_bench import demos
demos.bench_Lagrange()
The output of this code is::
#---------------------------------------------------------------------------
# Setting inputs of Lagrange polynomial functions: y=LAGRANGE(X,Y,x)
# where X is numpy.linspace(a,b,n+1), Y=fun(X) and x is random values on [a,b]
# n is the order of the Lagrange polynomial
# fun function is: lambda x: np.sin(x)
# [a,b]=[-1,1]
# X: (n+1,) numpy array
# Y: (n+1,) numpy array
# x: (m,) numpy array
# Error[i] computed with fun[i] output:
# lambda y: np.linalg.norm(y-fun(x),np.inf)
#---------------------------------------------------------------------------
# Benchmarking functions:
# fun[0], Lag: fc_bench.demos.Lagrange
# fun[1], lagint: fc_bench.demos.lagint
#---------------------------------------------------------------------------
# Comparative functions:
# comp[i-1,0], compares outputs of fun[0] and fun[i]
# lambda o1,o2: np.linalg.norm(o1-o2,np.inf)
# For each comparative function:
# - 1st input parameter is the output of fun[0]
# - 2nd input parameter is the output of fun[i]
#---------------------------------------------------------------------------
#date:2019-12-21_16-56-34
#nbruns:5
#numpy: <i8 <i8 f8 f8 f8 f8 f8
#format: {:>5} {:>5} {:6.3f} {:10.3e} {:9.3f} {:10.3e} {:9.3e}
#labels: m n Lag(s) Error[0] lagint(s) Error[1] comp[0]
100 3 0.014 1.218e-03 0.019 1.218e-03 2.220e-16
100 5 0.021 1.162e-05 0.028 1.162e-05 4.441e-16
100 7 0.030 6.999e-08 0.037 6.999e-08 3.331e-16
100 11 0.048 8.699e-13 0.056 8.689e-13 2.554e-15
500 3 0.068 1.218e-03 0.095 1.218e-03 3.331e-16
500 5 0.108 1.163e-05 0.144 1.163e-05 6.661e-16
500 7 0.152 7.004e-08 0.190 7.004e-08 8.882e-16
500 11 0.244 8.766e-13 0.282 8.766e-13 2.887e-15
