Update app.py

app.py

@@ -3,55 +3,83 @@ import joblib
 import pandas as pd
 import numpy as np
 
-# Load the trained pipeline
+# 1) Load your trained pipeline
 model = joblib.load("model.joblib")
 
-# Load team→specialty mapping from your dataset
+# 2) Load team→specialty mapping
 df = pd.read_csv("dataset.csv")
 team_specialty = df[['Team','Specialty']].drop_duplicates().set_index('Team')['Specialty'].to_dict()
 teams = list(team_specialty.keys())
 
-# Define the static task context
-new_task_base = {
-    'ProductType':             'Mothball',
-    'TaskType':                'Packaging',
-    'OrderQuantity':           100,
-    'DeadlineDays':            3,
-    'ExperienceYears':         5,
-    'AvgTaskTime_Minutes':     25.4,
-    'ErrorRate':               0.05,
-    'TrainingHours':           20.0,
-    'DayNumber':               2,
-    'ThroughputRate':          100 / 25.4,
-    'TimePressure':            100 / (3 * 25.4),
-    'PriorityLevel':           3
+# 3) Define several example task scenarios
+SCENARIOS = {
+    "Mothball → Packaging": {
+        'ProductType':         'Mothball',
+        'TaskType':            'Packaging',
+        'OrderQuantity':       100,
+        'DeadlineDays':        3,
+        'ExperienceYears':     5,
+        'AvgTaskTime_Minutes': 25.4,
+        'ErrorRate':           0.05,
+        'TrainingHours':       20.0,
+        'DayNumber':           2,
+        'ThroughputRate':      100 / 25.4,
+        'TimePressure':        100 / (3 * 25.4),
+        'PriorityLevel':       3
+    },
+    "Soap → Molding": {
+        'ProductType':         'Soap',
+        'TaskType':            'Molding',
+        'OrderQuantity':       250,
+        'DeadlineDays':        5,
+        'ExperienceYears':     4,
+        'AvgTaskTime_Minutes': 40.0,
+        'ErrorRate':           0.10,
+        'TrainingHours':       15.0,
+        'DayNumber':           1,
+        'ThroughputRate':      250 / 40.0,
+        'TimePressure':        250 / (5 * 40.0),
+        'PriorityLevel':       2
+    },
+    "Candle → QualityCheck": {
+        'ProductType':         'Candle',
+        'TaskType':            'QualityCheck',
+        'OrderQuantity':       150,
+        'DeadlineDays':        2,
+        'ExperienceYears':     6,
+        'AvgTaskTime_Minutes': 20.0,
+        'ErrorRate':           0.02,
+        'TrainingHours':       30.0,
+        'DayNumber':           3,
+        'ThroughputRate':      150 / 20.0,
+        'TimePressure':        150 / (2 * 20.0),
+        'PriorityLevel':       3
+    }
 }
 
-def generate_rankings():
-    # Build one row per team, filling in Team and Specialty
+def generate_rankings(scenario_name):
+    base = SCENARIOS[scenario_name]
     rows = []
     for team in teams:
-        row = new_task_base.copy()
-        row['Team'] = team
-        row['Specialty'] = team_specialty[team]
-        rows.append(row)
+        r = base.copy()
+        r['Team'] = team
+        r['Specialty'] = team_specialty[team]
+        rows.append(r)
     df_tasks = pd.DataFrame(rows)
-    
-    # Predict productivity for every team
-    preds = model.predict(df_tasks)
-    df_tasks['PredictedProductivity'] = np.round(preds, 2)
-    
-    # Return a sorted DataFrame
-    return df_tasks[['Team', 'PredictedProductivity']].sort_values(
-        'PredictedProductivity', ascending=False
+    df_tasks['PredictedProd'] = np.round(model.predict(df_tasks), 2)
+    return df_tasks[['Team', 'PredictedProd']].sort_values(
+        'PredictedProd', ascending=False
     ).reset_index(drop=True)
 
-# Gradio interface: single button → ranked table
 with gr.Blocks() as demo:
-    gr.Markdown("## Team Productivity Ranking for Today's Task")
+    gr.Markdown("## 🚀 Team Productivity Rankings")
+    scenario = gr.Dropdown(
+        choices=list(SCENARIOS.keys()),
+        label="Select a task scenario",
+        value=list(SCENARIOS.keys())[0]
+    )
     btn = gr.Button("Generate Rankings")
-    table = gr.Dataframe(headers=["Team", "PredictedProductivity"])
-    btn.click(fn=generate_rankings, inputs=None, outputs=table)
-    
-demo.launch()
+    table = gr.Dataframe(headers=["Team", "PredictedProd"])
+    btn.click(fn=generate_rankings, inputs=scenario, outputs=table)
 
+demo.launch()