import numpy as np
import pandas as pd
from catboost import CatBoostClassifier
from sklearn.model_selection import train_test_split
############best model fit#####################
data=pd.read_csv('./desktop/pancreatitis/dataset.csv')
X = data.drop(columns='hospital_expire_flag')
y = data['hospital_expire_flag']
X_train, X_test, y_train, y_test = train_test_split(X.astype(float), y.astype(float), test_size=0.20, random_state=0)
from sklearn.preprocessing import StandardScaler
sc = StandardScaler()
X_train = sc.fit_transform(X_train)
X_test = sc.transform(X_test)
method=CatBoostClassifier(n_estimators=208,learning_rate=0.009852052787531013,depth=6,subsample=0.6,l2_leaf_reg=6)
method.fit(X_train, y_train)
###########web-based prediction model############
import gradio as gr
import warnings
warnings.filterwarnings('ignore')
def PAN(Age,ALT,AST,BUN,DBP,INR,Lactate,SpO2,Temperature,WBC,Sepsis,CRRT,MV): 
    x = np.array([Age,ALT,AST,BUN,DBP,INR,Lactate,SpO2,Temperature,WBC,Sepsis,CRRT,MV])
    prediction = method.predict_proba(x.reshape(1, -1))[:,1]
    return prediction
outputs = gr.outputs.Textbox(label='Death Probability')
app = gr.Interface(fn=PAN, inputs=['number','number','number','number','number','number','number','number','number','number','number','number','number'], 
                   outputs=outputs,description="Death prediction with acute pancreatitis")
app.launch(share=True)