Update app.py

app.py

@@ -1,99 +1,162 @@
 import streamlit as st
 import pandas as pd
-import matplotlib.pyplot as plt
-
-# Load data
-data = pd.read_csv('customers.csv')
-
-
-# Function to display basic data information
-def display_data_info(data):
-    st.write("### Data Information")
-    st.write(data.info())
-    st.write("### Data Description")
-    st.write(data.describe())
-
-
-# Function to plot data
-def plot_data(data, column, chart_type):
-    if column == 'Net Sale':
-        data[column] = data[column].replace('[\$,]', '', regex=True).str.strip()
-        data[column] = pd.to_numeric(data[column], errors='coerce')
-    elif column == 'Cost':
-        data[column] = data[column].replace('[\$,]', '', regex=True).str.strip()
-        data[column] = pd.to_numeric(data[column], errors='coerce')
-    elif column == 'Gross Profit':
-        data[column] = data[column].replace('[\$,]', '', regex=True).str.strip()
-        data[column] = pd.to_numeric(data[column], errors='coerce')
-    elif column == ' AVG Inv. ':
-        data[column] = data[column].replace('[\$,]', '', regex=True).str.strip()
-        data[column] = pd.to_numeric(data[column], errors='coerce')
-
-    grouped_data = data.groupby('Mail ZIP')[column].sum()
-
-    plt.figure(figsize=(10, 6))
-    if chart_type == 'Bar Chart':
-        grouped_data.plot(kind='bar', color='skyblue')
-    elif chart_type == 'Scatter Plot':
-        plt.scatter(grouped_data.index, grouped_data.values, color='skyblue')
-    elif chart_type == 'Line Chart':
-        grouped_data.plot(kind='line', color='skyblue')
-    plt.title(f'{column} by ZIP Code')
-    plt.xlabel('ZIP Code')
-    plt.ylabel(column)
-    plt.xticks(rotation=45)
-    st.pyplot(plt)
-
-
-# Function to plot gross margin by ZIP Code
-def gross_margin_by_zip_code(data, chart_type):
-    margin_by_zip = data.groupby('Mail ZIP')['Gross Margin'].mean()
-    plt.figure(figsize=(10, 6))
-    if chart_type == 'Bar Chart':
-        margin_by_zip.plot(kind='bar', color='orange')
-    elif chart_type == 'Scatter Plot':
-        plt.scatter(margin_by_zip.index, margin_by_zip.values, color='orange')
-    elif chart_type == 'Line Chart':
-        margin_by_zip.plot(kind='line', color='orange')
-    plt.title('Gross Margin by ZIP Code')
-    plt.xlabel('ZIP Code')
-    plt.ylabel('Gross Margin (%)')
-    plt.xticks(rotation=45)
-    st.pyplot(plt)
-
-
-# Streamlit App
-def main():
-    st.title("Sales and Cost Dashboard")
-
-    # Sidebar options
-    st.sidebar.title("Options")
-    options = st.sidebar.radio(
-        "Select a Report", [
-                    'Data Information',
-                    'Net Sales by ZIP Code',
-                    'Cost by ZIP Code',
-                    'Gross Profit by ZIP Code',
-                    'Gross Margin by ZIP Code',
-                    'Average Invoice by ZIP Code'
-                    ]
-        )
-
-    chart_type = st.sidebar.selectbox("Select Chart Type", ['Bar Chart', 'Scatter Plot', 'Line Chart'])
-
-    if options == 'Data Information':
-        display_data_info(data)
-    elif options == 'Net Sales by ZIP Code':
-        plot_data(data, 'Net Sale', chart_type)
-    elif options == 'Cost by ZIP Code':
-        plot_data(data, 'Cost', chart_type)
-    elif options == 'Gross Profit by ZIP Code':
-        plot_data(data, 'Gross Profit', chart_type)
-    elif options == 'Gross Margin by ZIP Code':
-        gross_margin_by_zip_code(data, chart_type)
-    elif options == 'Average Invoice by ZIP Code':
-        plot_data(data, ' AVG Inv. ', chart_type)
-
-
-if __name__ == '__main__':
-    main()
+import plotly.express as px
+import plotly.graph_objects as go
+import openai
+
+# Add your OpenAI API key here
+openai.api_key = st.secret["OPEN_AI_KEY"]
+
+st.title("Transaction Analysis")
+
+# Load the data
+df_converted = pd.read_csv('payments_optimized.csv')
+
+# Convert CreatedDate to datetime
+df_converted['CreatedDate'] = pd.to_datetime(df_converted['CreatedDate'])
+
+# Function to analyze transaction data
+def analyze_transaction_data(df):
+    summary = {
+        "total_transactions": len(df),
+        "total_amount": df['Amount'].sum(),
+        "average_amount": df['Amount'].mean(),
+        "max_amount": df['Amount'].max(),
+        "min_amount": df['Amount'].min(),
+        "median_amount": df['Amount'].median(),
+        "transaction_trend": df.groupby(df['CreatedDate'].dt.to_period("M"))['Amount'].sum().tolist()
+    }
+    return summary
+
+# Function to get response from ChatGPT
+def get_chatgpt_response(prompt):
+    response = openai.ChatCompletion.create(
+        model="gpt-4o",
+        messages=[
+            {"role": "system", "content": "You are a helpful assistant."},
+            {"role": "user", "content": prompt}
+        ],
+        max_tokens=500
+    )
+    return response.choices[0].message['content'].strip()
+
+# Function to send report to ChatGPT
+def send_report(plot_title, data):
+    summary = analyze_transaction_data(data)
+    prompt = f"""
+    Transaction Data Insights:
+    1. Total Transactions: {summary['total_transactions']}
+    2. Total Amount Transacted: ${summary['total_amount']:.2f}
+    3. Average Transaction Amount: ${summary['average_amount']:.2f}
+    4. Maximum Transaction Amount: ${summary['max_amount']:.2f}
+    5. Minimum Transaction Amount: ${summary['min_amount']:.2f}
+    6. Median Transaction Amount: ${summary['median_amount']:.2f}
+
+    Provide strategic ideas based on this data.
+    """
+    response = get_chatgpt_response(prompt)
+    st.session_state['chat_with_gpt'] = f"Report for {plot_title} sent to ChatGPT. Data Summary: {summary}\n\nChatGPT Response: {response}"
+    st.experimental_rerun()
+
+# Plotting Transaction Amounts over Time with clickable dots
+st.subheader('Transaction Amounts Over Time')
+fig = px.scatter(
+    df_converted,
+    x='CreatedDate',
+    y='Amount',
+    custom_data=['CustomerName', 'PaymentMethod', 'Store', 'InvoiceType', 'TransactionType'],
+    title='Transaction Amounts Over Time'
+)
+
+fig.update_traces(
+    marker=dict(size=10),
+    selector=dict(mode='markers'),
+    hovertemplate=(
+        'Date: %{x}<br>'
+        'Amount: $%{y}<br>'
+        'Customer: %{customdata[0]}<br>'
+        'Payment Method: %{customdata[1]}<br>'
+        'Store: %{customdata[2]}<br>'
+        'Invoice Type: %{customdata[3]}<br>'
+        'Transaction Type: %{customdata[4]}<br>'
+    )
+)
+
+# Convert Plotly figure to a FigureWidget
+fig_widget = go.FigureWidget(fig)
+
+# Function to display click data
+def display_click_data(trace, points, state):
+    if points.point_inds:
+        point = points.point_inds[0]
+        selected_data = df_converted.iloc[point]
+        st.session_state['selected_point'] = selected_data.to_dict()
+        st.experimental_rerun()
+
+# Add click event handler
+fig_widget.data[0].on_click(display_click_data)
+
+# Display the plot
+st.plotly_chart(fig_widget)
+
+# Display details when a point is clicked
+if 'selected_point' in st.session_state:
+    st.write('Details for selected transaction:')
+    st.write(st.session_state['selected_point'])
+
+# Button to send report to ChatGPT
+if st.button("Executive Report - Transaction Amounts Over Time"):
+    send_report("Transaction Amounts Over Time", df_converted[
+        ['CreatedDate', 'Amount', 'CustomerName', 'PaymentMethod', 'Store', 'InvoiceType', 'TransactionType']]
+    )
+
+# Plotting Payment Method Preferences
+st.subheader('Payment Method Preferences')
+payment_method_counts_converted = df_converted['PaymentMethod'].value_counts().reset_index()
+payment_method_counts_converted.columns = ['PaymentMethod', 'Count']
+fig = px.bar(payment_method_counts_converted, x='PaymentMethod', y='Count', title='Payment Method Preferences')
+fig.update_layout(xaxis_title='Payment Method', yaxis_title='Number of Transactions')
+st.plotly_chart(fig)
+
+# Button to send report to ChatGPT
+if st.button("Executive Report - Payment Method Preferences"):
+    send_report("Payment Method Preferences", payment_method_counts_converted)
+
+# Calculate the total amount of payments for the most frequent customers
+top_customers = df_converted['CustomerName'].value_counts().index[:10]
+total_amounts = df_converted[df_converted['CustomerName'].isin(top_customers)].groupby('CustomerName', observed=True)[
+    'Amount'].sum().reset_index()
+
+# Plot the total payment amounts for the most frequent customers
+st.subheader('Total Payment Amounts for Most Frequent Customers')
+fig = px.bar(total_amounts, x='CustomerName', y='Amount', title='Total Payment Amounts for Most Frequent Customers')
+fig.update_layout(xaxis_title='Customer Name', yaxis_title='Total Transaction Amount ($)', yaxis_tickprefix='$')
+st.plotly_chart(fig)
+
+# Button to send report to ChatGPT
+if st.button("Executive Report - Total Payment Amounts for Most Frequent Customers"):
+    send_report("Total Payment Amounts for Most Frequent Customers", total_amounts)
+
+# Simulated chat with ChatGPT
+if 'chat_with_gpt' in st.session_state:
+    st.subheader("Chat with ChatGPT")
+    st.write(st.session_state['chat_with_gpt'])
+    user_input = st.text_input("Your question for ChatGPT:", key="chat_input")
+    if st.button("Send"):
+        if user_input:
+            summary = analyze_transaction_data(df_converted)
+            prompt = f"""
+            Transaction Data Insights:
+            1. Total Transactions: {summary['total_transactions']}
+            2. Total Amount Transacted: ${summary['total_amount']:.2f}
+            3. Average Transaction Amount: ${summary['average_amount']:.2f}
+            4. Maximum Transaction Amount: ${summary['max_amount']:.2f}
+            5. Minimum Transaction Amount: ${summary['min_amount']:.2f}
+            6. Median Transaction Amount: ${summary['median_amount']:.2f}
+
+            User Question: {user_input}
+            """
+            response = get_chatgpt_response(prompt)
+            st.write(f"ChatGPT Response: {response}")
+        else:
+            st.write("Please enter a question for ChatGPT.")