# import csv

import csv
import re

from flask import Flask, render_template, request, redirect,

url_for,session
from werkzeug.utils import secure_filename
import pymysql
db = pymysql.connect(host = 'localhost',user =
'root',password='',db='electricity')
cursor = db.cursor()

from sklearn.model_selection import train_test_split

from sklearn.ensemble import RandomForestRegressor
import xgboost as xgb
from sklearn.svm import SVR
from sklearn.metrics import
mean_absolute_error,mean_squared_error,r2_score
import os
import pandas as pd
app = Flask(__name__)
app.config['UPLOAD_FOLDER'] = 'uploads'
# app.secret_key = '..'

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

@app.route('/load data',methods = ["POST","GET"])

def load_data():
# if a>0:
if request.method == "POST":
f= request.files['file']
filetype = os.path.splitext(f.filename)[1]
#print("qew")
print(filetype)
#print('uuhj')
if filetype == '.csv':
mypath=os.path.join(app.config['UPLOAD_FOLDER'],f.filename)
#print('aaaaaa')
#print(mypath)
f.save(mypath)
print(mypath)
df = pd.read_csv(mypath)
type(df)
df.drop(['Unnamed: 0'],axis = 1,inplace = True)
print(df)
s = mypath
sql = "Truncate table tablename1"
cursor.execute(sql)
db.commit()
sql = "INSERT INTO tablename1 ("
for col in df.columns.values:
query = f"{col}, "
sql = sql + query
sql = sql[:-2] + ") values (" + ('%s, ' * 10)
sql = sql[:-2] + ")"
for row in df.iterrows():
 val = [str(row[1][i]) for i in range(len(row[1]))]
cursor.execute(sql, tuple(val))
db.commit()

return render_template('load data.html', msg='success')

elif filetype != '.csv':
return render_template('load data.html',msg = 'invalid')
return render_template('load data.html')
# else:
# return render_template('load.html',msg = 'fail')
# return render_template('load.html')
@app.route('/view data')
def view_data():
# if a>0:
# # print(mypath)
# file = os.listdir(app.config['UPLOAD_FOLDER'])
# data =
pd.read_csv(os.path.join(app.config['UPLOAD_FOLDER'],file[0]))
# data = pd.DataFrame(data)
# data .drop("Unnamed: 0",axis =1,inplace =True)
# global full_data
#
full_data1=pd.read_csv(os.path.join(app.config["UPLOAD_FOLDER"],myfile[0]))
data=pd.read_sql_query('SELECT * FROM tablename1', db)
data.drop(['id'], axis=1, inplace=True)
# data=clean_data(data)
#print(data)
# print(res.values.tolist())
return render_template('view data.html',msg = 'data',data =
data,col_name=data.columns.values,row_val=data.values.tolist())
# else:
# return render_template('view data.html',msg = 'fail')
@app.route('/model',methods = ["POST","GET"])
def model():
# if a>0:
if request.method == 'POST':
selected = int(request.form['selected'])
global df
testsize = int(request.form['testing'])
testsize = testsize/100
# global df
print('111')
filename = os.listdir(app.config['UPLOAD_FOLDER'])
# global df
df = pd.read_csv(os.path.join(app.config['UPLOAD_FOLDER'],
filename[0]))

df.drop("Unnamed: 0", axis=1, inplace=True)

X = df.drop(['priceactual'], axis=1)
y = df['priceactual']

global x_train, x_test, y_train, y_test

x_train, x_test, y_train, y_test = train_test_split(X, y,
test_size=testsize, random_state=10)
# global x_train, x_test, y_train, y_test
if (selected == 1):
rfr = RandomForestRegressor(n_estimators=50,max_depth=14)
model1 = rfr.fit(x_train,y_train)
pred1 = model1.predict(x_test)
 score = r2_score(y_test,pred1)
#print('aaa')
return render_template('model.html',msg = 'accuracy',result
= round(score,4),selected ='RANDOM FOREST REGRESSOR')
elif (selected == 2):
xgbr = xgb.XGBRegressor(learning_rate=0.4,n_estimators=200)
model2 = xgbr.fit(x_train,y_train)
pred2 = model2.predict(x_test)
score = r2_score(y_test,pred2)
return render_template('model.html',msg = 'accuracy',result
=round(score,4),selected = 'XGBOOST REGRESSOR')
elif (selected==3):
#print('hghfvh')
svr = SVR(kernel='rbf')
x_train = x_train[:15000]
y_train = y_train[:15000]
#print("u")
model3 = svr.fit(x_train[:15000],y_train[:15000])
#print('v')
pred3 = model3.predict(x_test[:15000])
#print('l')
score = r2_score(y_test[:15000],pred3)
#print('k')
return render_template('model.html',msg = 'accuracy',result
=round(score,4),selected = 'SUPPORT VECTOR REGRESSOR')
return render_template('model.html')
# else:
# return render_template('model.html', msg='fail')
@app.route('/registration',methods = ["POST","GET"])
def registration():
if request.method == "POST":

name = request.form['name']
email = request.form['email']
pwd = request.form['pwd']
cpwd = request.form['cpwd']
phno = request.form['phno']
# print('sss')
if pwd == cpwd:
global sql
sql = "select * from registration where name = '%s' and
email='%s'" %(name,email)
print(sql)
a = cursor.execute(sql)
# print(a)
if(a>0):
return render_template('registration.html',msg = 'invalid')
else:
sql = "insert into registration(name,email,pwd,phno) values
(%s,%s,%s,%s)"
print(sql)
# print('GHGH')
val= (name,email,pwd,phno)
# print('OIJO')
cursor.execute(sql,val)
# print('kjjh')
db.commit()
# print('lkjh')
return render_template('registration.html',msg = 'success')
 else:
return render_template('registration.html',msg='mismatch')
return render_template('registration.html')

@app.route('/login',methods=["POST","GET"])
def login():
if request.method == 'POST':
name = request.form['name']
pwd = request.form['pwd']
sql = "select * from registration where name = '%s' and pwd='%s'" %
(name,pwd)

a = cursor.execute(sql)

if a>0:
return render_template('index1.html')
else:
return render_template('login.html',msg = 'invalid')
return render_template('login.html')
@app.route('/admin',methods=["POST","GET"])
# def admin():
# if request.method == "POST":
# name = request.form['name']
# pwd = request.form['pwd']
# if name=='Admin' and pwd == 'admin123':
# return render_template('index1.html')
# else:
# return render_template('admin.html',msg = 'invalid')
# return render_template('admin.html')
@app.route('/index1')
def index():
return render_template('index1.html')
@app.route('/prediction',methods = ['POST','GET'])
def prediction():
# if a>0:
if request.method == "POST":
gfg = request.form['gfg']
gfhc = request.form['gfhc']
ghpsc = request.form['ghpsc']
ghwr = request.form['ghwr']
gor = request.form['gor']
gw = request.form['gw']
tlf = request.form['tlf']
tla = request.form['tla']
time = request.form['time']
values = [[float(gfg), float(gfhc), float(ghpsc), float(ghwr),
float(gor), float(gw), float(tlf), float(tla),float((time))]]
xgbr = xgb.XGBRegressor(learning_rate=0.4, n_estimators=200)
model = xgbr.fit(x_train, y_train)
df_pred = pd.DataFrame(values[0],
index=x_test.columns).transpose()
pred = model.predict(df_pred)
# score = r2_score(y_test, pred)
return render_template('prediction.html',msg = 'success',result
= pred)
return render_template('prediction.html')
# else:
# return render_template('prediction.html',msg = 'fail')
 # return render_template('predictions.html')
@app.route('/logout')
def logout():
return render_template('logout.html')
@app.route('/view users')
def view_users():
print('lkmn')
sql1 = "select * from registration"
cursor.execute(sql1)
data = cursor.fetchall()
print(data)
df = pd.read_sql_query(sql1,con=db)
g = re.split(',',data)
print(g)
d = pd.DataFrame(g,columns=['id','name','email','pwd','phno'])

# sql = cursor.execute(sql)
# print()
return render_template('view users.html',table = d)

if __name__ == ('__main__'):
app.run(debug=True)