"""
Detecting the location of a lighthouse in an idiomatic style.
"""

# 3.3.1 Arrange your import statements in a standard order
from dataclasses import dataclass
import math
import pickle
import random

import matplotlib
from matplotlib import pyplot as plt
import numpy

matplotlib.interactive(True)


# PEP-8 Formatting
# Use new-style classes
# Use dataclasses
@dataclass
class UFO:
    """
    A UFO that shoots death rays.
    """
    position: float = 2
    height: float = 1

    def emit_pulse(self):
        """

        Returns
        -------
        float
            Distance at which the pulse is detected.
        """
        # Shadow name of built-in
        direction = random.uniform(0, 2 * 3.14159265)
        #
        #        _U
        #       /d|h
        # +---+l--+
        # 0   1   2
        #
        # tan(d) = (2 - l) / h
        # h * tan(d) = 2 - l
        # l = 2 - h * tan(d)

        h = 2 - self.height * math.tan(direction)
        return direction, h


# PEP-8 Formatting
def log_posterior_location(height, prior_location, hs):
    """
    Get the log of the posterior of pos_shore.

    See Sivia-Skilling Data Analysis Bayesian Tutorial 2.38.

    Parameters
    ----------
    height: float
        Known height of the UFO.
    prior_location: array_like
        Locations sampled from the prior.
    hs: array_like
        Measured locations.
    """
    # 3.2.1 Follow the docstring conventions described in PEP-257

    priors2, hs2 = numpy.meshgrid(prior_location, hs)
    ph2 = numpy.log(height ** 2 + (hs2 - priors2) ** 2)
    po3 = - ph2.sum(axis=0)

    # Remove outdated code, use version control for that.
    print(po3)

    return po3


# 3.5.1 Use the if __name__ == '__main__' pattern to allow a file to be both imported and run directly
def main():
    ufo = UFO()

    priorsh = numpy.linspace(-2, 10, 100)

    hs = []
    metans = []
    metahs = []
    metaposts = []
    for n in range(6):
        hs = []
        for i in range(10**n):
            d, h = ufo.emit_pulse()
            hs.append(h)

        h_mean = sum(hs) / len(hs)
        log_post = log_posterior_location(ufo.height, priorsh, numpy.array(hs))
        log_postmax = numpy.argmax(log_post)
        h_post = priorsh[log_postmax]
        metans.append(n)
        metahs.append(h_mean)
        metaposts.append(h_post)
        print(n, h_mean, h_post)

    # 2.8.1 Use a context manager to ensure resources are properly managed
    with open("detections.pickle", 'wb') as f:
        pickle.dump(hs, f)

    # Use object-oriented matplotlib
    # See https://pbpython.com/effective-matplotlib.html
    fig, ax = plt.subplots()
    ax.scatter(metans, metahs)
    ax.scatter(metans, metaposts)
    ax.axhline(ufo.position)
    fig.show()

    fig2, ax2 = plt.subplots()
    ax2.hist(hs, range=(0, 5), bins=100)
    fig2.show()


# 3.5.1 Use the if __name__ == '__main__' pattern to allow a file to be both imported and run directly
if __name__ == '__main__':
    main()

# PEP-8: End with single empty line