Update pages/3_EDA_and_Feature_Engineering.py

pages/3_EDA_and_Feature_Engineering.py

@@ -1,177 +1,99 @@
 import streamlit as st
-import pandas as pd
 import seaborn as sns
 import matplotlib.pyplot as plt
-from sklearn.preprocessing import LabelEncoder
 
-st.title("Exploratory Data Analysis (EDA)")
+# Page Title
+st.title("Complete EDA Page")
 
-if 'df' in st.session_state:
-    df = st.session_state['df']  # Retrieve dataset
-    st.write("### Dataset Preview")
-    st.dataframe(df)
+st.markdown("""
+### Perform a Detailed Exploratory Data Analysis (EDA)
+This page uses the previously uploaded dataset and provides comprehensive analysis, including advanced visualizations and insights.
+---
+""")
 
+# Check if dataset exists in session state
+if 'df' in st.session_state:
+    data = st.session_state['df']
+    st.success("Dataset loaded successfully from previous upload.")
+
+    # Display Dataset Overview
+    st.write("### Dataset Overview")
+    st.write(data.head())
+
+    st.write("### General Information")
+    st.write(f"Number of Rows: {data.shape[0]}")
+    st.write(f"Number of Columns: {data.shape[1]}")
+    st.write("Column Names:", list(data.columns))
+
+    # Visualize Numeric Columns
+    numeric_columns = data.select_dtypes(include=['float64', 'int64']).columns
+    if len(numeric_columns) > 0:
+        st.write("### Numeric Column Analysis")
+
+        # Pairplot
+        st.write("#### Pairplot for Numeric Columns")
+        fig = sns.pairplot(data[numeric_columns], diag_kind='kde', palette='husl')
+        st.pyplot(fig)
+
+        # Correlation Heatmap
+        st.write("#### Correlation Heatmap")
+        corr_matrix = data[numeric_columns].corr()
+        fig, ax = plt.subplots(figsize=(10, 8))
+        sns.heatmap(corr_matrix, annot=True, cmap='coolwarm', ax=ax)
+        st.pyplot(fig)
+
+        # Histograms
+        st.write("#### Histograms")
+        for col in numeric_columns:
+            fig, ax = plt.subplots()
+            sns.histplot(data[col], kde=True, palette='crest', ax=ax)
+            ax.set_title(f'Histogram of {col}')
+            st.pyplot(fig)
+
+    else:
+        st.write("No numeric columns available for analysis.")
+
+    # Visualize Categorical Columns
+    categorical_columns = data.select_dtypes(include=['object', 'category']).columns
+    if len(categorical_columns) > 0:
+        st.write("### Categorical Column Analysis")
+        
+        # Frequency Count Plots
+        for col in categorical_columns:
+            st.write(f"#### Bar Plot for {col}")
+            fig, ax = plt.subplots()
+            sns.countplot(x=col, data=data, palette='viridis', ax=ax)
+            ax.set_title(f'Bar Plot of {col}')
+            st.pyplot(fig)
+
+    else:
+        st.write("No categorical columns available for analysis.")
+
+    # Advanced Analysis: Boxplots
+    if len(numeric_columns) > 0 and len(categorical_columns) > 0:
+        st.write("### Boxplot Analysis")
+        
+        selected_num_col = st.selectbox("Select Numeric Column for Boxplot", numeric_columns)
+        selected_cat_col = st.selectbox("Select Categorical Column for Boxplot", categorical_columns)
+
+        fig, ax = plt.subplots()
+        sns.boxplot(x=selected_cat_col, y=selected_num_col, data=data, palette='mako', ax=ax)
+        ax.set_title(f'Boxplot of {selected_num_col} by {selected_cat_col}')
+        st.pyplot(fig)
+
+    # Missing Values Heatmap
+    st.write("### Missing Values Analysis")
+    if data.isnull().values.any():
+        st.write("#### Missing Value Heatmap")
+        fig, ax = plt.subplots(figsize=(10, 6))
+        sns.heatmap(data.isnull(), cbar=False, cmap='viridis', ax=ax)
+        st.pyplot(fig)
+    else:
+        st.write("No missing values found in the dataset.")
+
+    # Summary Statistics
     st.write("### Summary Statistics")
-    st.write(df.describe())
-
-    # Rename columns
-    df = df.rename(columns={'ProductCategory': 'Category', 'ProductBrand': 'Brand', 'ProductPrice': 'Price'})
-
-    # Correct Price Column values (round off to 2 decimal places)
-    df['Price'] = df['Price'].apply(lambda x: round(x, 2))
-
-    st.markdown("Complete EDA + Feature Engineering")
-    st.markdown("## **Univariate Analysis**")
-
-    st.markdown("### PRODUCT CATEGORY")
-    st.write(fig, ax = plt.subplots(figsize=(10, 6))
-    sns.countplot(x='Category', data=df, palette='viridis', ax=ax)
-    ax.set_title("Product Category Distribution")
-    ax.set_xlabel("Product Category")
-    ax.set_ylabel("Count")
-    ax.tick_params(axis='x', rotation=45))
-    st.plt(fig)
-    st.markdown("""
-    **Insights:**
-    - 5 product categories observed.
-    - Highest frequency: Smart Phones and Laptops.
-    - Lowest frequency: Tablets and Headphones.
-    """)
-
-    st.markdown("### BRANDS")
-    fig, ax = plt.subplots(figsize=(10, 6))
-    sns.countplot(x='Brand', data=df, palette='cubehelix', ax=ax)
-    ax.set_title("Product Brand Distribution")
-    ax.set_xlabel("Product Brand")
-    ax.set_ylabel("Count")
-    ax.tick_params(axis='x', rotation=45)
-    st.plt(fig)
-    st.markdown("""
-    **Insights:**
-    - Samsung and HP have the highest frequencies.
-    - Sony, Apple, and other brands follow behind.
-    """)
-
-    st.markdown("### PRICE DISTRIBUTION")
-    fig, ax = plt.subplots(figsize=(10, 6))
-    sns.histplot(df['Price'], kde=True, color='orange', ax=ax)
-    ax.set_title("Product Price Distribution")
-    ax.set_xlabel("Product Price")
-    ax.set_ylabel("Count")
-    st.plt(fig)
-    st.markdown("""
-    **Insights:**
-    - Products span a wide price range, from near 0 to 3000.
-    - Significant concentration between 200 and 2500.
-    """)
-
-    st.markdown("### PRODUCT PRICE BINNING")
-    df['ProductPriceBucket'] = pd.cut(df['Price'], bins=[100, 500, 1000, 1500, 2000, 3000], labels=['Very Low', 'Low', 'Medium', 'High', 'Very High'])
-    fig, ax = plt.subplots(figsize=(10, 6))
-    sns.countplot(x='ProductPriceBucket', data=df, palette='icefire', ax=ax)
-    ax.set_title("Product Price Distribution")
-    ax.set_xlabel("Product Price Bucket")
-    ax.set_ylabel("Count")
-    ax.tick_params(axis='x', rotation=45)
-    st.plt(fig)
-    st.markdown("""
-    **Insights:**
-    - "Very High" price bucket has the highest concentration.
-    - Fewer products in the "Very Low" bucket.
-    - "Low", "Medium", and "High" have moderate distribution.
-    """)
-
-    st.markdown("### AGE DISTRIBUTION AND BINNING")
-    df['CustomerAgeGroup'] = pd.qcut(df['CustomerAge'], q=4, labels=['Young', 'Middle-aged', 'Mature', 'Senior'])
-    fig, axs = plt.subplots(1, 2, figsize=(15, 6))
-
-    sns.countplot(x='CustomerAgeGroup', data=df, ax=axs[0], palette='magma')
-    axs[0].set_title("Customer Age Group Distribution")
-    axs[0].set_xlabel("Customer Age Group")
-
-    sns.histplot(df['CustomerAge'], kde=True, ax=axs[1], color='teal')
-    axs[1].set_title("Customer Age Distribution")
-    axs[1].set_xlabel("Customer Age")
-
-    plt.tight_layout()
-    st.plt(fig)
-    st.markdown("""
-    **Insights:**
-    - Age groups are relatively evenly distributed.
-    - Slightly higher representation for "Young".
-    """)
-
-    st.markdown("### GENDER DISTRIBUTION")
-    fig, ax = plt.subplots(figsize=(10, 8))
-    df['CustomerGender'].value_counts().plot(kind='pie',
-                                             colors=['lightblue', 'lightpink'],
-                                             autopct='%1.1f%%',
-                                             startangle=90,
-                                             wedgeprops={'edgecolor': 'black'},
-                                             ax=ax)
-    ax.set_title("Customer Gender Distribution")
-    st.plt(fig)
-    st.markdown("""
-    **Insights:**
-    - Gender distribution is almost equal.
-    - Female: 50.9%, Male: 49.1%.
-    """)
-
-    st.markdown("### PURCHASE FREQUENCY DISTRIBUTION")
-    fig, ax = plt.subplots(figsize=(10, 6))
-    sns.histplot(df['PurchaseFrequency'], kde=True, color='purple', bins=30, ax=ax)
-    ax.set_title("Purchase Frequency Distribution")
-    ax.set_xlabel("Purchase Frequency")
-    ax.set_ylabel("Count")
-    st.plt(fig)
-    st.markdown("""
-    **Insights:**
-    - Purchase frequencies range from 1 to 19.
-    """)
-
-    st.markdown("### CUSTOMER SATISFACTION DISTRIBUTION")
-    fig, ax = plt.subplots(figsize=(10, 6))
-    sns.histplot(df['CustomerSatisfaction'], kde=True, color=sns.color_palette("crest", n_colors=1)[0], ax=ax)
-    ax.set_title("Customer Satisfaction Distribution")
-    ax.set_xlabel("Customer Satisfaction")
-    ax.set_ylabel("Count")
-    st.plt(fig)
-    st.markdown("""
-    **Insights:**
-    - Distinct peaks around integer values (1-5).
-    - Indicates customers tend to provide whole-number ratings.
-    """)
-
-    st.markdown("### PURCHASE INTENT DISTRIBUTION")
-    fig, ax = plt.subplots(figsize=(8, 6))
-    purchase_intent_counts = df['PurchaseIntent'].value_counts()
-    ax.pie(purchase_intent_counts,
-           labels=purchase_intent_counts.index,
-           colors=sns.color_palette("coolwarm", n_colors=len(purchase_intent_counts)),
-           autopct='%1.1f%%',
-           startangle=90,
-           wedgeprops={'edgecolor': 'black'})
-    ax.set_title("Purchase Intent Distribution")
-    st.plt(fig)
-    st.markdown("""
-    **Insights:**
-    - Binary classification problem (0: Not Purchase, 1: Purchase).
-    - 56.6% intent to purchase, 43.4% do not.
-    """)
-
-    st.markdown("### CORRELATION HEATMAP")
-    label_encoder = LabelEncoder()
-    df['CustomerGender'] = label_encoder.fit_transform(df['CustomerGender'])
-    df['Brand'] = label_encoder.fit_transform(df['Brand'])
-    df['Category'] = label_encoder.fit_transform(df['Category'])
-
-    corr_matrix = df.corr()
-    fig, ax = plt.subplots(figsize=(12, 8))
-    sns.heatmap(corr_matrix, annot=True, cmap='coolwarm', fmt='.2f', linewidths=0.5, ax=ax)
-    ax.set_title("Correlation Heatmap")
-    st.plt(fig)
-    st.markdown("""
-    **Insights:**
-    - Strong correlations can be observed between certain variables.
-    - Customer Satisfaction and Purchase Intent might have meaningful relationships.
-    """)
+    st.write(data.describe(include='all'))
+
+else:
+    st.error("No dataset found. Please upload a dataset on the previous page first.")