Update app.py

app.py

@@ -1,88 +1,82 @@
 import gradio as gr
-import pandas as pd
+import joblib
 import numpy as np
+import pandas as pd
 import matplotlib.pyplot as plt
-import joblib
+import random
 
 # Load model
-model = joblib.load("model.joblib")
+model = joblib.load("model.joblib")  # Ensure your model was saved as a Pipeline
 
-# Define example tasks
-example_tasks = {
-    # 🟢 High Productivity Task
-    "Candle → QualityCheck (High Productivity)": {
-        'quarter': 'Q1', 'department': 'quality', 'day': 'Tuesday',
-        'no_of_workers': 55, 'incentive': 3.5, 'idle_time': 0.1,
-        'idle_men': 2, 'smv': 20.0, 'month': 2, 'day_of_week': 1,
-        'is_weekend': 0, 'smv_per_worker': 20.0 / 55,
-        'effort_index': 20.0 + 3.5 + 1.0 - 0.1,
-        'log_wip': np.log1p(30), 'log_overtime': np.log1p(1.0),
-        'no_of_style_change': 0, 'targeted_productivity': 0.9
-    },
+# Load base data for teams
+df = pd.read_csv("data.csv")  # Used to get team names
+teams = sorted(df['Team'].unique())
 
-    # 🟡 Medium Productivity Task
-    "Soap → Molding (Moderate Productivity)": {
-        'quarter': 'Q2', 'department': 'sewing', 'day': 'Wednesday',
-        'no_of_workers': 48, 'incentive': 2.5, 'idle_time': 0.3,
-        'idle_men': 4, 'smv': 30.0, 'month': 5, 'day_of_week': 2,
-        'is_weekend': 0, 'smv_per_worker': 30.0 / 48,
-        'effort_index': 30.0 + 2.5 + 1.0 - 0.3,
-        'log_wip': np.log1p(50), 'log_overtime': np.log1p(1.0),
-        'no_of_style_change': 0, 'targeted_productivity': 0.75
-    },
+# Define possible randomized tasks
+def get_random_task():
+    tasks = [
+        {
+            'quarter': 'Q1', 'department': 'molding', 'day': 'Monday', 'no_of_workers': 40,
+            'incentive': 2.0, 'idle_time': 0.3, 'idle_men': 3, 'smv': 25.0, 'month': 3,
+            'day_of_week': 0, 'is_weekend': 0, 'log_wip': np.log1p(35), 'log_overtime': np.log1p(0.8),
+            'no_of_style_change': 0, 'targeted_productivity': 0.75
+        },
+        {
+            'quarter': 'Q3', 'department': 'packaging', 'day': 'Wednesday', 'no_of_workers': 52,
+            'incentive': 1.5, 'idle_time': 0.6, 'idle_men': 6, 'smv': 28.0, 'month': 8,
+            'day_of_week': 2, 'is_weekend': 0, 'log_wip': np.log1p(60), 'log_overtime': np.log1p(0.5),
+            'no_of_style_change': 1, 'targeted_productivity': 0.7
+        },
+        {
+            'quarter': 'Q4', 'department': 'quality', 'day': 'Friday', 'no_of_workers': 35,
+            'incentive': 3.0, 'idle_time': 0.1, 'idle_men': 1, 'smv': 32.0, 'month': 11,
+            'day_of_week': 4, 'is_weekend': 0, 'log_wip': np.log1p(20), 'log_overtime': np.log1p(1.2),
+            'no_of_style_change': 0, 'targeted_productivity': 0.85
+        }
+    ]
+    task = random.choice(tasks)
+    task['smv_per_worker'] = task['smv'] / task['no_of_workers']
+    task['effort_index'] = task['smv'] + task['incentive'] + 1.0 - task['idle_time']
+    return task
 
-    # 🔴 Low Productivity Task
-    "Mothball → Packaging (Low Productivity)": {
-        'quarter': 'Q4', 'department': 'finishing', 'day': 'Saturday',
-        'no_of_workers': 38, 'incentive': 1.0, 'idle_time': 1.5,
-        'idle_men': 10, 'smv': 35.0, 'month': 12, 'day_of_week': 5,
-        'is_weekend': 1, 'smv_per_worker': 35.0 / 38,
-        'effort_index': 35.0 + 1.0 + 1.0 - 1.5,
-        'log_wip': np.log1p(90), 'log_overtime': np.log1p(0.5),
-        'no_of_style_change': 2, 'targeted_productivity': 0.60
-    }
-}
-
-
-# Define team list
-teams = [f"team_{i}" for i in range(1, 13)]
-
-# Prediction function
-def predict_productivity(task_name):
-    base = example_tasks[task_name]
-    predictions = []
+# Function to predict productivity and return sorted chart and summary
+def predict():
+    task = get_random_task()
+    team_scores = []
 
     for team in teams:
-        row = base.copy()
-        row['Team'] = team
-        df_row = pd.DataFrame([row])
-        pred = model.predict(df_row)[0]
-        predictions.append((team, round(pred, 5)))
+        t = task.copy()
+        t['Team'] = team
+        task_df = pd.DataFrame([t])
+        pred = model.predict(task_df)[0]
+        team_scores.append((team, pred))
 
-    result_df = pd.DataFrame(predictions, columns=["Team", "Predicted Productivity"])
-    result_df = result_df.sort_values(by="Predicted Productivity", ascending=False)
+    df_scores = pd.DataFrame(team_scores, columns=["Team", "Predicted Productivity"])
+    df_scores = df_scores.sort_values(by="Predicted Productivity", ascending=False)
 
-    # Plot bar chart
-    fig, ax = plt.subplots(figsize=(10, 6))
-    ax.barh(result_df["Team"], result_df["Predicted Productivity"], color="skyblue")
+    # Plot
+    fig, ax = plt.subplots(figsize=(8, 6))
+    ax.barh(df_scores['Team'], df_scores['Predicted Productivity'], color='skyblue')
     ax.set_xlabel("Predicted Productivity")
-    ax.set_title(f"Team Ranking for: {task_name}")
+    ax.set_title("📊 Team Ranking for Random Task")
     ax.invert_yaxis()
     plt.tight_layout()
 
-    return result_df, fig
+    # Convert task to readable summary
+    task_summary = "\n".join([f"{k}: {v}" for k, v in task.items()])
+
+    return task_summary, fig
 
 # Gradio UI
 demo = gr.Interface(
-    fn=predict_productivity,
-    inputs=gr.Dropdown(choices=list(example_tasks.keys()), label="Select Task Scenario"),
+    fn=predict,
+    inputs=[],
     outputs=[
-        gr.Dataframe(label="🏆 Team Productivity Rankings:"),
-        gr.Plot(label="📈 Team Productivity Graph")
+        gr.Textbox(label="📌 Random Task Details"),
+        gr.Plot(label="📈 Team Productivity Rankings")
     ],
-    title="👷 Team Assignment Predictor",
-    description="Select a task to predict productivity scores for all teams. The best team will be at the top!"
+    title="🧠 Team Assignment Predictor",
+    description="Click Submit to generate a random task and rank all teams based on predicted productivity."
 )
 
-if __name__ == "__main__":
-    demo.launch()
+demo.launch()