markov_letters

markov_letters, an Octave code which counts the occurences of letter pairs in a text.

The text

  The quick brown fox jumps over the lazy dog.

The function reads the text, which will generally be stored in a file. It counts the occurrence of every possible pair of letters, creating a table(), a 27x27 array of frequencies. Entry table(1,3) counts the number of times the letter 'a' was immediately followed by the letter 'c', for instance. The 27th row and column keep track of the blank or non-alphabetic characters.

If you divide each row of the table by its sum, then this normalized value of table(i,j) can be regarded as the probability that a randomly selected occurrence of the i-th letter will be immediately followed by the j-th letter. This suggests various statistical tests on texts, and a means of simulating a text at a very low level.

Licensing:

The computer code and data files described and made available on this web page are distributed under the MIT license

Languages:

markov_letters is available in a MATLAB version and an Octave version.

Related Software and Data:

markov_letters_test

chrpak, an Octave code which works with characters and strings.

filum, an Octave code which can work with information in text files.

Source Code:

• bigram_frequency.m constructs a bigram frequency table from the text in a file;
• bigram_frequency_print.m prints a bigram frequency table;
• bigram_probability_print.m prints a bigram probability table;
• unigram_frequency.m constructs a unigram frequency table from the text in a file;
• unigram_frequency_print.m prints a unigram frequency table;