A6 part 2 utilities

a6_utils.py

#!/usr/bin/env python2
# -*- coding: utf-8 -*-
"""
Provides A6 part 2 utils
"""
import random


_CHARS_SPACE = ['A', 'Y', 'Z', ' ']
_CHARS = ['A', 'Y', 'Z']
_SYMBOLS = {
    'A': [1, 0, 1, 1, 1],
    'Y': [1, 1, 1, 0, 1, 0, 1, 1, 1, 0, 1, 1, 1],
    'Z': [1, 1, 1, 0, 1, 1, 1, 0, 1, 0, 1],
    '': [0, 0, 0],
    ' ': [0, 0, 0, 0, 0, 0, 0]
    }


def _create_evidence(string):
    '''Yields bits of an evidence vector from an input string'''
    for i, c in enumerate(string[:-1]):
        yield _SYMBOLS[c]
        if c != ' ' and string[i + 1] != ' ':
            yield _SYMBOLS['']

    yield _SYMBOLS[string[-1]]


def create_evidence(string):
    '''Returns an evidence vector from a string input

    Parameters
    ----------
    string : str
        Input string to convert to evidence list

    Returns
    -------
    List[int]
        A sequence of 1s and 0s corresponding to evidence bits
    '''
    evidence = []
    for l in _create_evidence(string):
        evidence.extend(l)
    return evidence


def _generate_string(length=None, max_len=10):
    '''Yields characters that form a valid HMM input string'''
    length = random.randint(1, max_len) if length is None else length
    if length < 3:
        for _ in xrange(length):
            yield random.choice(_CHARS)
    else:
        c = random.choice(_CHARS)
        yield c
        for _ in xrange(length - 2):
            c = random.choice(_CHARS) if c == ' ' else random.choice(_CHARS_SPACE)
            yield c
        yield random.choice(_CHARS)


def generate_string(length=None, max_len=10):
    '''Returns a random HMM input string

    Parameters
    ----------
    length : int, optional
       Length of generated string. If None, is randomly chosen [1, `max_len`]
    max_len : int, optional
        Maximum length of random string if `length` is None

    Returns
    -------
    str
        Valid HMM input string
    '''
    return ''.join(_generate_string(length, max_len))


def merge_dicts(*dicts):
    '''Merges multiple dicts into a single dict. Useful for combining probability dicts

    Parameters
    ----------
    *dicts
        Variable number of `dict` objects to merge

    Returns
    -------
    dict
        A single merged dict
    '''
    merged = dict()
    for d in dicts:
        merged.update(d)
    return merged


if __name__ == '__main__':
    '''Example usage'''
    from hmm_submission import part_2_b

    # TODO: create variables: states, prior_probs, transition_probs, emission_probs

    input_string = generate_string()  # or set this manually, e.g. 'AYZ'
    evidence_vector = create_evidence(input_string)

    output_string, _ = part_2_b(evidence_vector, states, prior_probs, transition_probs, emission_probs)

    assert input_string == output_string