graph.py

import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
from preprocess import split_sentences, split_words

def graph_sentence_lengths(files, window=20):
    plt.figure(figsize=(12, 6))
    
    for file in files:
        with open(file.name, 'r', encoding='utf-8') as f:
            text = f.read()
        
        sentences = split_sentences(text)
        lengths = []
        for sent in sentences:
            words = split_words(sent)
            lengths.append(len(words))
        
        n = len(lengths)
        if n > 1:
            smoothed = []
            half = window // 2
            for i in range(n):
                start = max(0, i - half)
                end = min(n, i + half + 1)
                smoothed.append(np.median(lengths[start:end]))
            
            x_normalized = np.linspace(0, 1, n)
            plt.plot(x_normalized, smoothed, label=file.name.split('/')[-1], linewidth=1.5)
    
    plt.xlabel('Относительная позиция в тексте')
    plt.ylabel('Длина предложения (сглажено)')
    plt.title('Распределение длин предложений по тексту')
    plt.legend(loc='upper right', fontsize=8)
    plt.grid(True, alpha=0.3)
    return plt.gcf()