RandProfileHistogramUniform

class poisson_approval.RandProfileHistogramUniform(n_bins, orders=None, d_order_fixed_share=None, **kwargs)[source]

A random factory of histogram profiles (ProfileHistogram) following the uniform distribution.

Parameters:
  • n_bins (int) – The number of bins in each histogram. Cf. ProfileHistogram.
  • orders (list, optional) – These orders will have a variable share. They can be rankings, e.g. 'abc', or weak orders, e.g. 'a~b>c'. Default: all rankings.
  • d_order_fixed_share (dict, optional.) – A dictionary. For each entry order: fixed_share, this order will have at least this fixed share. The total must be lower or equal to 1.
  • kwargs (keyword arguments) – These additional arguments will be passed directly to ProfileHistogram.

Examples

Basic usage:

>>> initialize_random_seeds()
>>> rand_profile = RandProfileHistogramUniform(n_bins=3)
>>> profile = rand_profile()
>>> print(profile)
<abc: 0.4236547993389047 [0.43758721 0.2083069  0.35410589], acb: 0.12122838365799216 [0.891773   0.07188976 0.03633724], bac: 0.0039303209304278885 [0.38344152 0.40828352 0.20827496], bca: 0.05394987214431912 [0.52889492 0.03914964 0.43195544], cab: 0.1124259903007756 [0.07103606 0.85456058 0.07440336], cba: 0.2848106336275805 [0.0202184 0.0669109 0.9128707]> (Condorcet winner: a)

Using the optional parameters:

>>> rand_profile = RandProfileHistogramUniform(
...     n_bins=3, orders=['abc', 'a~b>c'], d_order_fixed_share={'b>a~c': 0.5},
...     voting_rule=PLURALITY)
>>> profile = rand_profile()
>>> print(profile)
<abc: 0.416309922773969 [0.77815675 0.0918554  0.12998785], a~b>c: 0.083690077226031, b>a~c: 0.5> (Condorcet winner: b) (Plurality)