Code fragments of linearRegression.py

• import pickle
• from random import randrange
• import matplotlib
• import matplotlib.pyplot as plt
• import seaborn as sns
• sns.set()
• import numpy as np
• from sklearn.utils.validation import check_is_fitted
• from datetime import datetime
• import mysql.connector
• from mysql.connector import errorcode
• import os
• def is_fitted(estimator):
• try:
• check_is_fitted(estimator)
• return True
• except: return False
• HOST = os.environ['DB_HOST']
• DB_NAME = os.environ['DB_NAME']
• USERNAME = os.environ['DB_USERNAME']
• PASSWORD = os.environ['DB_PASSWORD']
• TABLE = 'lr'
• def get_data():
• x = []
• y = []
• #connect
• try:
• cnx = mysql.connector.connect(user=USERNAME, password=PASSWORD, host=HOST, database=DB_NAME)
• print("Successfully connected to MySQL server")
• cursor = cnx.cursor()
• print("Successfully create a cursor for the connection")
• try:
• cursor.execute("USE {}".format(DB_NAME))
• cursor.execute("select x from {}".format(TABLE))
• x = np.array(cursor.fetchall(), dtype=np.float64).flatten()
• print("x is fetched")
• cursor.execute("select y from {}".format(TABLE))
• y = np.array(cursor.fetchall(), dtype=np.float64).flatten()
• print("y is fetched")
• except mysql.connector.Error as err: print(err)
• except mysql.connector.Error as err: if err.errno == errorcode.ER_ACCESS_DENIED_ERROR: print("Something is wrong with your user name or password") elif err.errno == errorcode.ER_BAD_DB_ERROR: print("Database does not exist") else: print(err)
• cursor.close()
• print('Cursor is closed')
• cnx.close()
• print('Connection is closed')
• return x,y
• x,y = get_data()
• filename = 'model.sav'
• loaded_model = pickle.load(open(filename, 'rb'))
• fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(15,5))
• ax1.title.set_text('Before Run')
• ax2.title.set_text('After Run')
• ax1.scatter(x[:, np.newaxis], y[:], c='blue')
• ax2.scatter(x[:, np.newaxis], y[:], c='blue')
• #Is it first run?
• if(is_fitted(loaded_model)):
• xfit = np.linspace(0, 10, 1000)
• yfitIncr1 = loaded_model.predict(xfit[:, np.newaxis])
• ax1.plot(xfit, yfitIncr1)
• else:
• ax1.plot([],[])
• #visit Station 1 first
• for _ in range(25):
• loaded_model.partial_fit(x[:, np.newaxis], y[:])
• pickle.dump(loaded_model, open(filename, 'wb'))
• xfit = np.linspace(0, 10, 1000)
• yfitIncr1 = loaded_model.predict(xfit[:, np.newaxis])
• ax2.plot(xfit, yfitIncr1)
• now = datetime.now()
• current_time = now.strftime("%H:%M:%S")
• print(current_time)
• fig.savefig(current_time+'.png')
• #plt.savefig('images/final.png')
• plt.close()

Graph