import streamlit as st import pandas as pd import plotly.express as px import matplotlib.pyplot as plt import numpy as np # Page configuration st.set_page_config(page_title="Customer Insights App", page_icon=":bar_chart:") # Load CSV files df = pd.read_csv("df_clean.csv") nombres_proveedores = pd.read_csv("nombres_proveedores.csv", sep=';') euros_proveedor = pd.read_csv("euros_proveedor.csv", sep=',') nombres_proveedores['codigo'] = nombres_proveedores['codigo'].astype(str) euros_proveedor['CLIENTE'] = euros_proveedor['CLIENTE'].astype(str) # Ignore the last two columns df = df.iloc[:, :-2] # Ensure customer code is a string df['CLIENTE'] = df['CLIENTE'].astype(str) # Function to get supplier name def get_supplier_name(code): name = nombres_proveedores[nombres_proveedores['codigo'] == code]['nombre'].values return name[0] if len(name) > 0 else code # Function to create radar chart def radar_chart(categories, values, amounts, title): # Number of variables N = len(categories) # Calculate angles for each point angles = [n / float(N) * 2 * np.pi for n in range(N)] angles += angles[:1] # Initialize the plot fig, ax = plt.subplots(figsize=(12, 12), subplot_kw=dict(projection='polar')) # Normalize values and amounts max_value = max(values) normalized_values = [v / max_value for v in values] total_amount = sum(amounts) normalized_amounts = [a / total_amount for a in amounts] # Draw polygon for units and fill it normalized_values += normalized_values[:1] ax.plot(angles, normalized_values, 'o-', linewidth=2, color='#FF69B4', label='% Units') ax.fill(angles, normalized_values, alpha=0.25, color='#FF69B4') # Draw polygon for amounts and fill it normalized_amounts += normalized_amounts[:1] ax.plot(angles, normalized_amounts, 'o-', linewidth=2, color='#4B0082', label='% Spend') ax.fill(angles, normalized_amounts, alpha=0.25, color='#4B0082') # Set axes ax.set_xticks(angles[:-1]) ax.set_xticklabels(categories, size=8, wrap=True) ax.set_ylim(0, max(max(normalized_values), max(normalized_amounts)) * 1.1) # Draw reference circles circles = np.linspace(0, 1, 5) for circle in circles: ax.plot(angles, [circle]*len(angles), '--', color='gray', alpha=0.3, linewidth=0.5) # Remove radial labels and chart borders ax.set_yticklabels([]) ax.spines['polar'].set_visible(False) # Add title and legend plt.title(title, size=16, y=1.1) plt.legend(loc='upper right', bbox_to_anchor=(1.3, 1.1)) return fig # Main page design st.title("Welcome to Customer Insights App") st.markdown(""" This app helps businesses analyze customer behaviors and provide personalized recommendations based on purchase history. Use the tools below to dive deeper into your customer data. """) # Navigation menu page = st.selectbox("Select the tool you want to use", ["", "Customer Analysis", "Customer Recommendations"]) # Home Page if page == "": st.markdown("## Welcome to the Customer Insights App") st.write("Use the dropdown menu to navigate between the different sections.") # Customer Analysis Page elif page == "Customer Analysis": st.title("Customer Analysis") st.markdown(""" Use the tools below to explore your customer data. """) # Customer filter field partial_code = st.text_input("Enter part of Customer Code (or leave empty to see all)") # Filter customer options that match the partial code if partial_code: filtered_customers = df[df['CLIENTE'].str.contains(partial_code)] else: filtered_customers = df # Create a list of filtered customers for the selectbox customer_list = filtered_customers['CLIENTE'].unique() # Customer selection with filtered autocomplete customer_code = st.selectbox("Select Customer Code", customer_list) if customer_code: # Filter data for the selected customer customer_data = df[df["CLIENTE"] == customer_code] customer_euros = euros_proveedor[euros_proveedor["CLIENTE"] == customer_code] if not customer_data.empty and not customer_euros.empty: st.write(f"### Analysis for Customer {customer_code}") # Define purchase threshold purchase_threshold = 0 # Get all manufacturers the customer bought from (ignore the customer column) all_manufacturers = customer_data.iloc[:, 1:].T[customer_data.iloc[:, 1:].T[customer_data.index[0]] > purchase_threshold] # Sort manufacturers by value in descending order all_manufacturers = all_manufacturers.sort_values(by=customer_data.index[0], ascending=False) # Prepare values and manufacturers values = all_manufacturers[customer_data.index[0]].values.tolist() manufacturers = [get_supplier_name(m) for m in all_manufacturers.index.tolist()] # Get amounts in euros amounts = [] for m in all_manufacturers.index.tolist(): if m in customer_euros.columns: amounts.append(customer_euros[m].values[0]) else: amounts.append(0) # If there are fewer than 3 manufacturers, add a third one with value 0 if len(manufacturers) < 3: manufacturers.append("Other") values.append(0) amounts.append(0) # Display the results for each manufacturer st.write(f"### Results for {len(manufacturers)} manufacturers (sorted):") for manufacturer, value, amount in zip(manufacturers, values, amounts): st.write(f"{manufacturer} = {value:.4f} units, €{amount:.2f}") # Create and display the radar chart fig = radar_chart(manufacturers, values, amounts, f'Radar Chart for {len(manufacturers)} Manufacturers of Customer {customer_code}') st.pyplot(fig) # Customer sales 2021-2024 (if data exists) if 'VENTA_2021' in df.columns and 'VENTA_2022' in df.columns and 'VENTA_2023' in df.columns and 'VENTA_2024' in df.columns: years = ['2021', '2022', '2023', '2024'] sales_columns = ['VENTA_2021', 'VENTA_2022', 'VENTA_2023', 'VENTA_2024'] customer_sales = customer_data[sales_columns].values[0] fig_sales = px.line(x=years, y=customer_sales, markers=True, title=f'Sales Over the Years for Customer {customer_code}') fig_sales.update_layout(xaxis_title="Year", yaxis_title="Sales") st.plotly_chart(fig_sales) else: st.warning("Sales data for 2021-2024 not available.") else: st.warning(f"No data found for customer {customer_code}. Please check the code.") # Customer Recommendations Page elif page == "Customer Recommendations": st.title("Customer Recommendations") st.markdown(""" Get tailored recommendations for your customers based on their purchasing history. """) # Customer filter field partial_code = st.text_input("Enter part of Customer Code for Recommendations (or leave empty to see all)") # Filter customer options that match the partial code if partial_code: filtered_customers = df[df['CLIENTE'].str.contains(partial_code)] else: filtered_customers = df # Create a list of filtered customers for the selectbox customer_list = filtered_customers['CLIENTE'].unique() # Customer selection with filtered autocomplete customer_code = st.selectbox("Select Customer Code for Recommendations", customer_list) if customer_code: customer_data = df[df["CLIENTE"] == customer_code] if not customer_data.empty: # Show selected customer's purchase history st.write(f"### Purchase History for Customer {customer_code}") st.write(customer_data) # Generate recommendations (placeholder) st.write(f"### Recommended Products for Customer {customer_code}") # You can replace this with the logic of the recommendation model st.write("Product A, Product B, Product C") else: st.warning(f"No data found for customer {customer_code}. Please check the code.")