Update app.py

app.py

@@ -18,10 +18,10 @@ def safe_convert(value, default, min_val, max_val):
     except (TypeError, ValueError):
         return default  # Use default if conversion fails
 
-# Create the main function for server
+# Create the main function for the model
 def main_func(Department, ChainScale, SupportiveGM, Merit, LearningDevelopment, WorkEnvironment, Engagement, WellBeing):
     
-    # ChainScale mapping
+    # These mappings are EXAMPLES only, not used in the model
     ChainScale_mapping = {
         'Luxury': 1,
         'Upper Midscale': 2,
@@ -29,10 +29,7 @@ def main_func(Department, ChainScale, SupportiveGM, Merit, LearningDevelopment,
         'Upscale': 4,
         'Independent': 5,
     }
-    default_ChainScale = 4  
-    ChainScale_value = ChainScale_mapping.get(ChainScale, default_ChainScale)
-
-    # Department mapping
+    
     department_mapping = {
         "Guest Services": 1,
         "Food and Beverage": 2,
@@ -40,9 +37,8 @@ def main_func(Department, ChainScale, SupportiveGM, Merit, LearningDevelopment,
         "Front Office Operations": 4,
         "Guest Activities": 5,
     }
-    default_department = 5  
-    department_value = department_mapping.get(Department, default_department)
 
+    # Convert inputs to safe numeric values
     LearningDevelopment = safe_convert(LearningDevelopment, 3.0, 1, 5)
     SupportiveGM = safe_convert(SupportiveGM, 3.0, 1, 5)
     Merit = safe_convert(Merit, 3.0, 1, 5)
@@ -50,9 +46,8 @@ def main_func(Department, ChainScale, SupportiveGM, Merit, LearningDevelopment,
     Engagement = safe_convert(Engagement, 3.0, 1, 5)
     WellBeing = safe_convert(WellBeing, 3.0, 1, 5)
 
+    # Only include model-relevant features
     new_row = pd.DataFrame({
-        'Department': [int(department_value)],  
-        'ChainScale': [int(ChainScale_value)],          
         'SupportiveGM': [SupportiveGM],
         'Merit': [Merit],
         'LearningDevelopment': [LearningDevelopment],
@@ -61,9 +56,9 @@ def main_func(Department, ChainScale, SupportiveGM, Merit, LearningDevelopment,
         'WellBeing': [WellBeing]
     }).astype(float)
 
+    # Predict probabilities
     prob = loaded_model.predict_proba(new_row) 
     
-    # Ensure probabilities return correctly
     if prob.shape[1] == 2:
         leave_prob = float(prob[0][0])
         stay_prob = float(prob[0][1])
@@ -71,6 +66,7 @@ def main_func(Department, ChainScale, SupportiveGM, Merit, LearningDevelopment,
         leave_prob = float(prob[0])
         stay_prob = 1 - leave_prob
 
+    # Generate SHAP values
     shap_values = explainer(new_row)
 
     fig, ax = plt.subplots(figsize=(8, 4))
@@ -88,7 +84,7 @@ def main_func(Department, ChainScale, SupportiveGM, Merit, LearningDevelopment,
 # Create the UI
 title = "**Mod 3 Team 5: Employee Turnover Predictor & Interpreter**"
 description1 = """
-This app predicts whether an employee intends to stay or leave based on satisfaction factors and department.
+This app predicts whether an employee intends to stay or leave based on satisfaction factors.
 """
 
 description2 = """
@@ -106,12 +102,12 @@ with gr.Blocks(title=title) as demo:
         with gr.Column():
             Department = gr.Radio(
                 ["Guest Services", "Food and Beverage", "Housekeeping", "Front Office Operations", "Guest Activities"],
-                label="Department", 
+                label="Department (Example Only)", 
                 value="Guest Services"
             )
             ChainScale = gr.Dropdown(
                 ["Luxury", "Upper Midscale", "Upper Upscale", "Upscale", "Independent"],
-                label="ChainScale", 
+                label="ChainScale (Example Only)", 
                 value="Upper Upscale"
             )
             SupportiveGM = gr.Slider(
@@ -166,4 +162,3 @@ with gr.Blocks(title=title) as demo:
     )
 
 demo.launch()
-