UnCodeGL package

Submodules

UnCodeGL.code module

class UnCodeGL.code.Code(code_dictionary)

Bases: object

Core class for Coding and Decoding. It has integrated core metrics to it.

To create a new Code you need to pass the rule for the code as a dictionary.

arity

An integer representing the code arity

code_alphabet

A set containing the code alphabet

source_alphabet

A set containing the source alphabet

decode(sentence, get_any=True)

Decodes a given sentence, in case multiple outputs are possible it returns any of them if the get_any flag is True.

Parameters

  • sentence – Sentence to be decoded. It must compel with the source alphabet.

  • get_any – True by default. It gives a string when multiple answers are possible. When False it gives the answer in a set of possible answers.

Returns

A string with the decoded sentence

encode(sentence)

“Encodes a given sentence

get_C_n()

Returns C_n related to the uniquely decodable algorithm.

Returns

List of set of words.

get_arity()

Returns arity of the code

Returns

Integer value of the arity

get_code_trie()

Returns a Trie data structure from the words in the code

get_word_length(source)

Gets the average word lenght of the code for a given Source

Parameters

source – A source of type Source

Returns

A float with the value of the word lenght of the code for the given source.

is_prefix()

Uses a Trie data structure to determine if the code is prefix

Returns

A boolean True if the code is prefix, False otherwise

is_uniquely_decodable()

Returns boolean specifying whether or not is uniquely decodable

For the code to be uniquely decodable means that the function code that takes a string from a the source and codifies it is inyective.

Returns

True if the code is uniquely decodable, False otherwise

static kraft_inequality_metric(words_length, arity)

Calculates Kraft Inequality for given word lengths

Measures Kraft Inequality Metric defined as

\[\sum_{i} r^{-l_i}\]

where r means arity, and l sub i the length of the ith word

Also returns if an instantaneous code can be made with said words

Typical Usage:

>>> metric, can_be_inst = kraft_inequality_metric(words_length=[4,4,3,6],arity=2)
>>> print(metric, can_be_inst)
0.265625 True
Parameters

  • words_length – An list with the length of the words

  • arity – Arity of the code

Returns

metric , can_be_instantaneous. Kraft Inequality metric. A boolean refering to whether or not a code with this word lenghts and arity exists. This is dependant on the metric being less than 1

static make_huffman_code(source, code_alphabet={'0', '1'})

Generates an optimal Huffman Code

Parameters

  • source – A source of type Source

  • code_alphabet – Iterable with the alphabet. The items should be of type str

Returns

Optimal huffman code for the given source and with arity of the given code_alphabet

UnCodeGL.code.validate_dictionary(code_dictionary)

Validates that a given dictionary is good to use as a code

UnCodeGL.source module

class UnCodeGL.source.Source(*args, **kwargs)

Bases: object

Class for Source. Contains a dictionary with the probabilities of each simbol it can generate

dictionary

Dictionary of probabilities. The keys are the source simbols and the values the probabilities

Initialization can be done through dictionary or base_string but not both

Can be initiated through a given dictionary that contains the simbols and the probabilities that each symbol occurs.

It can also be initiated with a base string that will count appearances and obtain probabilities.

Either dictionary or base_string must be passed, not both

Parameters

  • dictionary – A dictionary with the simbols and probabilities. The sum of the probabilities must be 1.

  • base_string – A string of the usual output of the source. It will be assumed that is a memoryless source

get_alphabet()

Returns list with simbols of the source

get_entropy()

Calculates entropy of the source

Entropy is given by

\[-\sum_{i} p(x_i)\log(p(x_i))\]
Returns

Float with the value of the entropy.

get_simbol_prob()

Returns probability for a simbol

UnCodeGL.source.make_dictionary_from_str(base_string)

Makes a dictionary with probabilities normalized

UnCodeGL.source.verify_dictionary(dictionary)

Verify the sum of the probabilites is 1 to an epsilon

Module contents