analysis_modules.py

# analysis_modules.py

import pandas as pd
import plotly.express as px
from statsmodels.tsa.seasonal import seasonal_decompose
from statsmodels.tsa.stattools import adfuller
from sklearn.cluster import KMeans
from wordcloud import WordCloud
import matplotlib.pyplot as plt
import io
import base64

# --- Time-Series Module ---
def analyze_time_series(df: pd.DataFrame, date_col: str, value_col: str):
    """Performs time-series decomposition and stationarity testing."""
    df[date_col] = pd.to_datetime(df[date_col])
    ts_df = df.set_index(date_col)[value_col].dropna()
    
    # Decomposition
    decomposition = seasonal_decompose(ts_df, model='additive', period=12) # Assuming monthly data
    fig_decomp = px.line(pd.DataFrame({'trend': decomposition.trend, 'seasonal': decomposition.seasonal, 'residual': decomposition.resid}),
                         title=f"Time-Series Decomposition of {value_col}")
                         
    # Stationarity Test (ADF)
    adf_result = adfuller(ts_df)
    adf_md = f"""
    ### Stationarity Analysis (ADF Test)
    - **Test Statistic:** `{adf_result[0]:.4f}`
    - **p-value:** `{adf_result[1]:.4f}`
    - **Conclusion:** The series is likely **{'stationary' if adf_result[1] < 0.05 else 'non-stationary'}**.
    """
    return fig_decomp, adf_md

# --- Text Analysis Module ---
def generate_word_cloud(df: pd.DataFrame, text_col: str):
    """Generates a word cloud from a text column."""
    text = ' '.join(df[text_col].dropna().astype(str))
    wordcloud = WordCloud(width=800, height=400, background_color='white').generate(text)
    
    # Convert matplotlib plot to a data URI for Gradio
    buf = io.BytesIO()
    wordcloud.to_image().save(buf, format='png')
    img_str = "data:image/png;base64," + base64.b64encode(buf.getvalue()).decode('utf-8')
    return img_str

# --- Clustering Module ---
def perform_clustering(df: pd.DataFrame, numeric_cols: list, n_clusters: int = 4):
    """Performs K-Means clustering and returns a scatter plot."""
    cluster_data = df[numeric_cols].dropna()
    kmeans = KMeans(n_clusters=n_clusters, random_state=42, n_init='auto').fit(cluster_data)
    cluster_data['Cluster'] = kmeans.labels_.astype(str)
    
    # For visualization, we'll use the first two numeric columns
    fig_cluster = px.scatter(cluster_data, x=numeric_cols[0], y=numeric_cols[1], color='Cluster',
                             title=f"K-Means Clustering (k={n_clusters})")
    return fig_cluster