pystylometry 1.3.0

Comprehensive Python package for stylometric analysis

pystylometry

Stylometric analysis and authorship attribution for Python. 50+ metrics across 11 modules, from vocabulary diversity to AI-generation detection.

Install

pip install pystylometry              # Core (lexical metrics)
pip install pystylometry[all]         # Everything

Individual extras

pip install pystylometry[readability]   # Readability formulas (pronouncing, spaCy)
pip install pystylometry[syntactic]     # POS/parse analysis (spaCy)
pip install pystylometry[authorship]    # Attribution methods
pip install pystylometry[ngrams]        # N-gram entropy
pip install pystylometry[viz]           # Matplotlib visualizations

Usage

from pystylometry.lexical import compute_mtld, compute_yule
from pystylometry.readability import compute_flesch

result = compute_mtld(text)
print(result.mtld_average)       # 72.4

result = compute_flesch(text)
print(result.reading_ease)       # 65.2
print(result.grade_level)        # 8.1

Every function returns a typed dataclass with the score, components, and metadata -- never a bare float.

Unified API

from pystylometry import analyze

results = analyze(text, lexical=True, readability=True, syntactic=True)

Style Drift Detection

Detect authorship changes, spliced content, and AI-generated text within a single document.

from pystylometry.consistency import compute_kilgarriff_drift

result = compute_kilgarriff_drift(document)
print(result.pattern)             # "sudden_spike"
print(result.pattern_confidence)  # 0.71
print(result.max_location)        # Window 23 -- the splice point

CLI

pystylometry-drift manuscript.txt --window-size=500 --stride=250
pystylometry-viewer report.html

Modules

Module	Metrics	Description
lexical	TTR, MTLD, Yule's K/I, Hapax, MATTR, VocD-D, HD-D, MSTTR, function words, word frequency	Vocabulary diversity and richness
readability	Flesch, Flesch-Kincaid, SMOG, Gunning Fog, Coleman-Liau, ARI, Dale-Chall, Fry, FORCAST, Linsear Write, Powers-Sumner-Kearl	Grade-level and difficulty scoring
syntactic	POS ratios, sentence types, parse tree depth, clausal density, passive voice, T-units, dependency distance	Sentence and parse structure (requires spaCy)
authorship	Burrows' Delta, Cosine Delta, Zeta, Kilgarriff chi-squared, MinMax, John's Delta, NCD	Author attribution and text comparison
stylistic	Contractions, hedges, intensifiers, modals, punctuation, vocabulary overlap (Jaccard/Dice/Cosine/KL), cohesion, genre/register	Style markers and text similarity
character	Letter frequencies, digit/uppercase ratios, special characters, whitespace	Character-level fingerprinting
ngrams	Word/character/POS n-grams, Shannon entropy, skipgrams	N-gram profiles and entropy
dialect	British/American classification, spelling/grammar/vocabulary markers, markedness	Regional dialect detection
consistency	Sliding-window chi-squared drift, pattern classification	Intra-document style analysis
prosody	Syllable stress, rhythm regularity	Prose rhythm (requires spaCy)
viz	Timeline, scatter, report (PNG + interactive HTML)	Drift detection visualization

Development

git clone https://github.com/craigtrim/pystylometry && cd pystylometry
pip install -e ".[dev,all]"
make test       # 1022 tests
make lint       # ruff + mypy
make all        # lint + test + build

License

MIT

Author

Craig Trim -- [e-mail]