app.py

import numpy as np
from flask import Flask, request, jsonify, render_template
import joblib
from sklearn.model_selection import train_test_split
from xgboost import XGBRegressor
import warnings
warnings.filterwarnings('ignore')

app = Flask(__name__)
lr_loaded = joblib.load('./static/lr.joblib')
xg_loaded = joblib.load('./static/xg.joblib')
etr_loaded = joblib.load('./static/etr.joblib')
gbr_loaded = joblib.load('./static/gbr.joblib')
br_loaded = joblib.load('./static/br.joblib')
rf_loaded = joblib.load('./static/rf.joblib')

@app.route('/')
def home():
    return render_template('index.html')

@app.route('/predict',methods=['POST'])
def predict():
    '''
    For rendering results on HTML GUI
    '''
    int_features=[]
    for x in request.form.values():
        try:
            int_features.append(int(float(x)))
        except ValueError:
            # Handle the case where conversion is not possible
            print(f"Skipping invalid value: {x}")

    # Now int_features contains the successfully converted integer values
    print(int_features)
    # int_features = [int(x) for x in request.form.values()]
    mo= int_features[0]
    prediction=[]
    output=0
    if mo==1:
        int_features= int_features[1:]
        final_features = np.array(int_features)
        pred_lr = lr_loaded.predict([final_features])
        pred_xg = xg_loaded.predict(final_features.reshape(1,-1))
        prediction= (pred_lr+pred_xg)/2

    elif mo==2:
        int_features= int_features[1:]
        final_features = [np.array(int_features)]
        pred_lr = lr_loaded.predict(final_features)
        pred_rf = rf_loaded.predict(final_features)
        prediction= (pred_lr+pred_rf)/2
    elif mo==3:
        int_features= int_features[1:]
        final_features = [np.array(int_features)]
        pred_lr = lr_loaded.predict(final_features)
        pred_gbr = gbr_loaded.predict(final_features)
        prediction= (pred_lr+pred_gbr)/2
    elif mo==4:
        int_features= int_features[1:]
        final_features = [np.array(int_features)]
        pred_br = br_loaded.predict(final_features)
        pred_gbr = gbr_loaded.predict(final_features)
        prediction= (pred_br+pred_gbr)/2
    elif mo==5:
        int_features= int_features[1:]
        final_features = [np.array(int_features)]
        pred_br = br_loaded.predict(final_features)
        pred_etr = etr_loaded.predict(final_features)
        prediction= (pred_br+pred_etr)/2
    else:
        int_features= int_features[1:]
        final_features = [np.array(int_features)]
        pred_lr = lr_loaded.predict(final_features)
        pred_br = br_loaded.predict(final_features)
        prediction= (pred_lr+pred_br)/2
        
# dels":["lr&xg","lr&rf","lr&gbr","br&gbr","br&etr","lr&br"],

    output = round(prediction[0], 2)
    print(output)

    return render_template('temp.html', prediction_text='The predicted load carrying capacity of the sample is {}kN'.format(output))

if __name__ == "__main__":
    app.run(host='0.0.0.0',port=5000)