gps_soln.py

"""This file contains code for use with "Think Bayes",
by Allen B. Downey, available from greenteapress.com

Copyright 2014 Allen B. Downey
License: GNU GPLv3 http://www.gnu.org/licenses/gpl.html
"""

from __future__ import print_function, division

import numpy
import thinkbayes2
import thinkplot

from itertools import product


class Gps(thinkbayes2.Suite, thinkbayes2.Joint):
    """Represents hypotheses about your location in the field."""

    def Likelihood(self, data, hypo):
        """Computes the likelihood of the data under the hypothesis.

        hypo: 
        data: 
        """
        std = 30
        meanx, meany = hypo
        x, y = data
        like = thinkbayes2.EvalNormalPdf(x, meanx, std)
        like *= thinkbayes2.EvalNormalPdf(y, meany, std)
        return like


def main():
    coords = numpy.linspace(-100, 100, 101)
    joint = Gps(product(coords, coords))

    joint.Update((51, -15))
    joint.Update((48, 90))

    pairs = [(11.903060613102866, 19.79168669735705),
             (77.10743601503178, 39.87062906535289),
             (80.16596823095534, -12.797927542984425),
             (67.38157493119053, 83.52841028148538),
             (89.43965206875271, 20.52141889230797),
             (58.794021026248245, 30.23054016065644),
             (2.5844401241265302, 51.012041625783766),
             (45.58108994142448, 3.5718287379754585)]

    joint.UpdateSet(pairs)

    thinkplot.PrePlot(2)
    pdfx = joint.marginal(0)
    pdfy = joint.marginal(1)
    thinkplot.Pdf(pdfx, label='posterior x')
    thinkplot.Pdf(pdfy, label='posterior y')
    thinkplot.Show()

    print(pdfx.mean(), pdfx.Std())
    print(pdfy.mean(), pdfy.Std())


if __name__ == '__main__':
    main()