import joblib
from sklearn.preprocessing import StandardScaler
import gradio as gr

input_module1 = gr.inputs.Textbox(label = "age (integer): ")
input_module2 = gr.inputs.Textbox(label = "sex (binary): ")
input_module3 = gr.inputs.Textbox(label = "chest pain type: ")
input_module4 = gr.inputs.Textbox(label = "resting blood pressure: ")
input_module5 = gr.inputs.Textbox(label = "serum cholestoral in mg/dl: ")
input_module6 = gr.inputs.Textbox(label = "fasting blood sugar > 120 mg/dl : ")
input_module7 = gr.inputs.Textbox(label = "resting electrocardiographic results: ")
input_module8 = gr.inputs.Textbox(label = "maximum heart rate achieved: ")
input_module9 = gr.inputs.Textbox(label = "exercise induced angina: ")
input_module10 = gr.inputs.Textbox(label = "oldpeak: ")
input_module11 = gr.inputs.Textbox(label = "slope of the peak exercise ST segment : ")
input_module12 = gr.inputs.Textbox(label = "number of major vessels colored by flourosopy ")
input_module13 = gr.inputs.Textbox(label = "thal: ")

input_module_2 = gr.inputs.Dropdown(choices=['Logistic Regression', 'Support Vector Machine', 'Decision Tree', 'Random Forest', 'Artificial Neural Networks', 'KNN', 'Linear Discriminant Analysis'], label = "ML Models Dropdown")

output_module1 = gr.outputs.Label(label = "Predicted Class")

lr_model = joblib.load('lr_model.joblib')
svc_model = joblib.load('svc_model.joblib')
dt_model = joblib.load('dt_model.joblib')
rf_model = joblib.load('rf_model.joblib')
mlp_model = joblib.load('mlp_model.joblib')
knn_model = joblib.load('knn_model.joblib')
lda_model = joblib.load('lda_model.joblib')

def pred(input1, input2, input3, input4, input5, input6, input7, input8, input9, input10, input11, input12, input13, input14):
  heart_data = pd.read_csv("heart.csv", header = None, sep=' ')
  heart_data.columns = ['age', 'sex', 'chest_pain_type', 'resting_blood_pressure', 'serum_cholestoral', 'fasting_blood_sugar', 'resting_ecg_results', 'max_heart_rate_achieved', 'exercise_induced_angina', 'oldpeak', 'slope_peak_exercise_st_segment', 'num_major_vessels_colored_by_flourosopy', 'thal', 'target_variable']
  
  X = heart_data.drop('target_variable', axis=1)
  y = heart_data['target_variable']

  scaler = StandardScaler()
  X = scaler.fit_transform(X)
  X_new = scaler.transform(X)

  if input14 == "Logistic Regression":
    preds = lr_model.predict(X_new)
  elif input14 == "Support Vector Machine":
    preds = svc_model.predict(X_new)
  elif input14 == "Decision Tree":
    preds = dt_model.predict(X_new)
  elif input14 == "Random Forest":
    preds = rf_model.predict(X_new)
  elif input14 == "Artificial Neural Networks":
    preds = mlp_model.predict(X_new)
  elif input14 == "KNN":
    preds = knn_model.predict(X_new)
  elif input14 == "Linear Discriminant Analysis":
    preds = lda_model.predict(X_new)

  if preds[0] == 1:
    results = "No heart disease"
  elif preds[0] == 2:
    results = "Heart disease"

  return results

gr.Interface(fn=pred, 
             inputs=[input_module1,input_module2,input_module3,input_module4,input_module5,input_module6,input_module7,input_module8,input_module9,input_module10,input_module11,input_module12,input_module13,input_module_2],
             outputs=output_module1).launch(debug = True)