LinguaF provides an easy access for researchers and developers to methods of quantitative language analysis, such as: readability, complexity, diversity, and other descriptive statistics.

documents = [
    "Pain and suffering are always inevitable for a large intelligence and a deep heart. The really great men must, I think, have great sadness on earth.",
    "To go wrong in one's own way is better than to go right in someone else's.",
    "The darker the night, the brighter the stars, The deeper the grief, the closer is God!"
]

The following descriptive statistics are supported (descriptive_statistics.py module):

• Number of characters char_count
• Number of letters letter_count
• Number of punctuation characters punctuation_count
• Number of digits digit_count
• Number of syllables syllable_count
• Number of sentences sentence_count
• Number of n-syllable words number_of_n_syllable_words
• Average syllables per word avg_syllable_per_word
• Average word length avg_word_length
• Average sentence length avg_sentence_length
• Average words per sentence avg_words_per_sentence

Additional methods:

• Get lexical items (nouns, adjectives, verbs, adverbs) get_lexical_items
• Get n-grams get_ngrams
• Get sentences get_sentences
• Get words get_words
• Tokenize tokenize
• Remove punctuation remove_punctuation
• Remove digits remove_digits

Example:

from linguaf import descriptive_statistics as ds


ds.words_per_sentence(documents)
# Output: 15

The following syntactical complexity metrics are supported (syntactical_complexity.py module):

• Mean Dependency Distance (MDD) mean_dependency_distance

Example:

from linguaf import syntactical_complexity as sc


sc.mean_dependency_distance(documents)
# Output: 2.307306255835668

The following lexical diversity metrics are supported (lexical_diversity.py module):

• Lexical Density (LD) lexical_density
• Type Token Ratio (TTR) type_token_ratio
• Herdan's Constant or Log Type Token Ratio (LogTTR) log_type_token_ratio
• Summer's Index summer_index
• Root Type Token Ratio (RootTTR) root_type_token_ratio

Example:

from linguaf import lexical_diversity as ld


ld.log_type_token_ratio(documents)
# Output: 94.03574963462502

The following readability metrics are supported (readability.py module):

• Flesch Reading Ease (FRE) flesch_reading_ease
• Flesch-Kincaid Grade (FKG) flesch_kincaid_grade
• Automated Readability Index (ARI) automated_readability_index
• Simple Automated Readability Index (sARI) automated_readability_index_simple
• Coleman's Readability Score coleman_readability
• Easy Listening Score easy_listening

Example:

from linguaf import readability as r


r.flesch_kincaid_grade(documents)
# Output: 4.813333333333336

pip install linguaf

git clone https://github.com/Perevalov/LinguaF.git
cd LinguaF
pip install .

At the moment, library supports English and Russian languages for all the methods.

TBD