app.py

import gradio as gr
import pandas as pd
import numpy as np
import plotly.express as px
import time
import os

# -------------------------------
# Utility Functions & Data Loading
# -------------------------------
def load_data():
    """
    Load the dataset.
    Update the filename/path based on your uploaded data.
    """
    filename = "data.csv"  # change this to your dataset filename if needed
    if os.path.exists(filename):
        try:
            data = pd.read_csv(filename)
        except Exception as e:
            print("Error reading file:", e)
            data = pd.DataFrame()
    else:
        # In case the file is not found, create a dummy dataframe for demonstration
        data = pd.DataFrame({
            "Date": pd.date_range(start="2023-01-01", periods=12, freq="M"),
            "Branch": np.random.choice(["North", "South", "East", "West"], 12),
            "Transactions": np.random.randint(100, 1000, 12),
            "Revenue": np.random.uniform(1000, 10000, 12)
        })
    return data

# Global load of data (simulate caching for performance)
DATA = load_data()

# -------------------------------
# Data Modeling and Aggregation
# -------------------------------
def data_modelling():
    """
    Demonstrates data modeling by performing built-in aggregations.
    This function computes aggregates such as total transactions and average revenue per branch.
    """
    if DATA.empty:
        return "No data loaded."
    
    # Built-in aggregations using pandas (simulate DAX-like functions)
    agg_data = DATA.groupby("Branch").agg({
        "Transactions": "sum",
        "Revenue": "mean"
    }).reset_index()
    
    # Rename columns to simulate DAX-like output formatting
    agg_data.rename(columns={
        "Transactions": "Total Transactions",
        "Revenue": "Average Revenue"
    }, inplace=True)
    
    return agg_data

# -------------------------------
# Dashboard / Data Visualization
# -------------------------------
def create_visualization():
    """
    Create interactive charts using Plotly.
    Generates two charts:
    1. A bar chart of total transactions per branch.
    2. A line chart of revenue over time.
    """
    if DATA.empty:
        return "No data available for visualization."
    
    # Bar Chart: Total Transactions by Branch
    agg = DATA.groupby("Branch")["Transactions"].sum().reset_index()
    bar_fig = px.bar(agg, x="Branch", y="Transactions", title="Total Transactions by Branch",
                     labels={"Transactions": "Total Transactions"})
    
    # Line Chart: Revenue over Time (if Date column exists)
    if "Date" in DATA.columns:
        # Ensure date is in datetime format
        DATA["Date"] = pd.to_datetime(DATA["Date"], errors='coerce')
        line_fig = px.line(DATA.sort_values("Date"), x="Date", y="Revenue", title="Revenue Over Time",
                           labels={"Revenue": "Revenue ($)"})
    else:
        line_fig = None
    
    return bar_fig, line_fig

# -------------------------------
# Simulated DAX Simulator
# -------------------------------
def dax_simulator(dax_command: str):
    """
    A simulated DAX command processor.
    This function pretends to interpret a DAX command and returns a pseudo-code or explanation.
    In a real project, you could integrate a generative AI API to translate DAX to pandas code.
    """
    dax_command = dax_command.strip().lower()
    if "sum" in dax_command:
        response = "DAX SUM detected. In pandas, you can use: df['Column'].sum()"
    elif "average" in dax_command or "mean" in dax_command:
        response = "DAX AVERAGE detected. In pandas, you can use: df['Column'].mean()"
    elif "count" in dax_command:
        response = "DAX COUNT detected. In pandas, you can use: df['Column'].count()"
    else:
        response = "DAX command not recognized. Please use SUM, AVERAGE/MEAN, or COUNT as an example."
    
    return response

# -------------------------------
# Generative AI Insights Generator
# -------------------------------
def generate_insights(query: str):
    """
    This function simulates generative AI insights.
    In a production system, you might integrate an API (e.g., OpenAI's GPT) to generate detailed insights.
    Here, we simulate with a simple response based on keywords.
    """
    query_lower = query.strip().lower()
    if "performance" in query_lower:
        insights = ("The data model shows strong performance with efficient aggregations. "
                    "Consider further analysis on revenue trends across different branches for optimization.")
    elif "trend" in query_lower:
        insights = ("The revenue trend over time indicates seasonality. Look into the impact of quarterly marketing "
                    "campaigns to better understand these patterns.")
    elif "optimization" in query_lower:
        insights = ("Data model performance can be improved by indexing key columns and caching frequently used aggregates. "
                    "Ensure your queries are optimized for mobile and desktop views.")
    else:
        insights = ("Based on the provided query, further investigation is recommended. "
                    "Analyze branch performance, transaction trends, and revenue patterns for actionable insights.")
    
    # Simulate processing time
    time.sleep(1)
    return insights

# -------------------------------
# Gradio Interface Layout
# -------------------------------
with gr.Blocks(css=".gradio-container { max-width: 100%; }") as demo:
    gr.Markdown("# 365 Business Bank Dashboard")
    gr.Markdown("This application showcases features such as Data Modelling, Dashboard Creation, Mobile Specific layout, Data Visualization, Builtin Aggregations, and simulated DAX commands. It also includes a generative AI insights module.")
    
    with gr.Tabs():
        # --- Data Overview Tab ---
        with gr.Tab("Data Overview"):
            gr.Markdown("### Data Model & Aggregations")
            agg_table = gr.Dataframe(value=data_modelling(), label="Aggregated Data by Branch", interactive=False)
            gr.Markdown("The above table shows total transactions and average revenue per branch computed using built-in aggregations.")
        
        # --- Visualization Tab ---
        with gr.Tab("Visualization"):
            gr.Markdown("### Interactive Charts")
            bar_chart = gr.Plot(label="Bar Chart: Total Transactions by Branch")
            line_chart = gr.Plot(label="Line Chart: Revenue Over Time")
            
            def update_charts():
                bar_fig, line_fig = create_visualization()
                return bar_fig, line_fig
            
            update_btn = gr.Button("Update Charts")
            update_btn.click(fn=update_charts, inputs=[], outputs=[bar_chart, line_chart])
        
        # --- DAX Simulator Tab ---
        with gr.Tab("DAX Simulator"):
            gr.Markdown("### Simulated DAX to pandas Conversion")
            dax_input = gr.Textbox(lines=2, placeholder="Enter a DAX command (e.g., SUM(Transactions))", label="DAX Command")
            dax_output = gr.Textbox(label="Pandas Equivalent Explanation")
            dax_input.submit(fn=dax_simulator, inputs=dax_input, outputs=dax_output)
            gr.Markdown("Enter common DAX functions (SUM, AVERAGE, COUNT) to see how they translate to pandas.")
        
        # --- AI Insights Generator Tab ---
        with gr.Tab("Insights Generator"):
            gr.Markdown("### Generate Business Insights")
            insight_input = gr.Textbox(lines=2, placeholder="Ask a question about business performance or trends", label="Your Query")
            insight_output = gr.Textbox(label="Generated Insights")
            insight_input.submit(fn=generate_insights, inputs=insight_input, outputs=insight_output)
            gr.Markdown("This simulated generative AI tool provides actionable insights based on your query. For a production system, integrate a dedicated AI service.")
    
    # Optional: A footer section
    gr.Markdown("© 2025 365 Business Bank Dashboard - Mobile and Desktop Optimized")

if __name__ == "__main__":
    demo.launch()