analyze.py

import streamlit as st
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt

def analyze_page(df):
    st.header("🎨 Advanced Data Analysis")
    st.write("Explore and analyze your dataset with colorful and interactive visualizations.")

    # Set a colorful seaborn theme
    sns.set_theme(style="whitegrid", context="talk", palette="colorblind")

    # Data Summary
    st.subheader("📊 Data Summary")
    st.write("Get an overview of the dataset's basic statistics.")
    st.write(df.describe())

    # Option to display full summary statistics or specific statistics
    if st.checkbox("Show more summary statistics"):
        st.write(df.describe(include='all'))

    # Missing Data Visualization
    st.subheader("🔍 Missing Data Analysis")
    missing_data = df.isnull().sum()
    st.write("Number of missing values in each column:")
    st.write(missing_data)

    if missing_data.sum() > 0:
        fig, ax = plt.subplots(figsize=(14, 8))
        sns.heatmap(df.isnull(), cbar=False, cmap='magma', ax=ax, linewidths=0.5)
        ax.set_title('Missing Data Heatmap', fontsize=18, fontweight='bold', color='darkviolet')
        ax.set_xlabel('Columns', fontsize=14, color='darkorange')
        ax.set_ylabel('Rows', fontsize=14, color='darkorange')
        st.pyplot(fig)
    else:
        st.write("No missing data in the dataset.")

    # Data Types
    st.subheader("🧩 Data Types")
    st.write("Data types of each column:")
    st.write(df.dtypes)

    # Correlation Analysis
    st.subheader("🔗 Correlation Analysis")
    st.write("Explore correlations between numeric variables.")
    
    # Filter numeric columns
    numeric_df = df.select_dtypes(include=['number'])
    
    if not numeric_df.empty:
        corr = numeric_df.corr()

        fig, ax = plt.subplots(figsize=(16, 12))
        sns.heatmap(corr, annot=True, fmt='.2f', cmap='coolwarm', ax=ax, linewidths=1, linecolor='white')
        ax.set_title('Correlation Heatmap', fontsize=18, fontweight='bold', color='crimson')
        st.pyplot(fig)
    else:
        st.write("No numeric columns available for correlation analysis.")

    # Interactive Column Analysis: Select Columns
    st.subheader("🔍 Interactive Column Analysis")
    st.write("Select columns to analyze their summary statistics, data types, and unique values.")
    
    selected_columns = st.multiselect("Select columns to analyze", df.columns)
    
    if selected_columns:
        st.write("Summary Statistics:")
        st.write(df[selected_columns].describe())
        st.write("Data Types:")
        st.write(df[selected_columns].dtypes)
        
        for col in selected_columns:
            st.write(f"Unique values in {col}:")
            st.write(df[col].unique())
            
            if pd.api.types.is_numeric_dtype(df[col]):
                st.write(f"Distribution of {col}:")
                bins = st.slider(f"Select number of bins for {col}", min_value=10, max_value=100, value=30)
                kde = st.checkbox(f"Add KDE for {col}", value=True)
                fig, ax = plt.subplots(figsize=(12, 8))
                sns.histplot(df[col], bins=bins, kde=kde, color='teal', edgecolor='black', ax=ax)
                ax.set_title(f'Distribution of {col}', fontsize=18, fontweight='bold', color='royalblue')
                ax.set_xlabel(f'{col}', fontsize=14)
                ax.set_ylabel('Frequency', fontsize=14)
                st.pyplot(fig)

    # Distribution of Numeric Columns (Overall)
    st.subheader("📈 Distribution of Numeric Columns")
    numeric_columns = numeric_df.columns
    dist_column = st.selectbox("Select a column to view its distribution", numeric_columns)
    
    if dist_column:
        bins = st.slider(f"Select number of bins for {dist_column}", min_value=10, max_value=100, value=30)
        kde = st.checkbox(f"Add KDE for {dist_column}", value=True)
        fig, ax = plt.subplots(figsize=(12, 8))
        sns.histplot(df[dist_column], bins=bins, kde=kde, color='mediumseagreen', edgecolor='black', ax=ax)
        ax.set_title(f'Distribution of {dist_column}', fontsize=18, fontweight='bold', color='royalblue')
        ax.set_xlabel(f'{dist_column}', fontsize=14)
        ax.set_ylabel('Frequency', fontsize=14)
        st.pyplot(fig)

    # Additional Analysis Plots

    # Box Plot
    st.subheader("📦 Box Plot")
    box_column = st.selectbox("Select a numeric column for Box Plot", numeric_columns)
    if box_column:
        fig, ax = plt.subplots(figsize=(12, 8))
        sns.boxplot(x=df[box_column], color='salmon', ax=ax)
        ax.set_title(f'Box Plot of {box_column}', fontsize=18, fontweight='bold', color='firebrick')
        ax.set_xlabel(f'{box_column}', fontsize=14)
        ax.set_ylabel('Values', fontsize=14)
        st.pyplot(fig)

    # Violin Plot
    st.subheader("🎻 Violin Plot")
    violin_column = st.selectbox("Select a numeric column for Violin Plot", numeric_columns)
    if violin_column:
        fig, ax = plt.subplots(figsize=(12, 8))
        sns.violinplot(x=df[violin_column], color='skyblue', ax=ax)
        ax.set_title(f'Violin Plot of {violin_column}', fontsize=18, fontweight='bold', color='cornflowerblue')
        ax.set_xlabel(f'{violin_column}', fontsize=14)
        ax.set_ylabel('Density', fontsize=14)
        st.pyplot(fig)

    # Count Plot
    st.subheader("📊 Count Plot")
    categorical_columns = df.select_dtypes(include=['object']).columns
    count_column = st.selectbox("Select a categorical column for Count Plot", categorical_columns)
    if count_column:
        fig, ax = plt.subplots(figsize=(12, 8))
        sns.countplot(x=df[count_column], palette='pastel', ax=ax)
        ax.set_title(f'Count Plot of {count_column}', fontsize=18, fontweight='bold', color='brown')
        ax.set_xlabel(f'{count_column}', fontsize=14)
        ax.set_ylabel('Count', fontsize=14)
        st.pyplot(fig)

    # Bar Plot
    st.subheader("📊 Bar Plot")
    bar_column = st.selectbox("Select a categorical column for Bar Plot", categorical_columns)
    if bar_column:
        bar_data = df[bar_column].value_counts()
        fig, ax = plt.subplots(figsize=(14, 10))
        sns.barplot(x=bar_data.index, y=bar_data.values, palette='viridis', ax=ax)
        ax.set_title(f'Bar Plot of {bar_column}', fontsize=18, fontweight='bold', color='darkviolet')
        ax.set_xlabel(f'{bar_column}', fontsize=14)
        ax.set_ylabel('Count', fontsize=14)
        st.pyplot(fig)

    # Custom Interactive Widgets
    st.sidebar.header("🛩️ Advanced Filters")
    min_value, max_value = st.sidebar.slider(
        "Select the range of Age",
        min_value=10,
        max_value=100,
        value=(10,100)
    )

    filtered_df = df[(df[numeric_columns[0]] >= min_value) & (df[numeric_columns[0]] <= max_value)]
    st.write(f"Filtered dataset based on the selected range: {min_value} - {max_value}")
    st.write(filtered_df)

    if st.sidebar.button("Reset Filters"):
        st.experimental_rerun()