- Python Python Python Python Python Python Python Python Python Python Python Python Python Python Python Python Python 1 (763.4 KiB) Viewed 19 times
def ReadCSV(filename): with open(filename) as csv_file: csv_reader = csv.reader(csv_file,delimiter=',') x_mylist = [] y_mylist = [] for row in csv_reader: x_mylist.append(float(row[0]) ) y_mylist.append(float(row[1]) ) return x_mylist, y_mylist
# ******************_x, _y = ReadCSV('ParabolaWithNoiseOutlier.csv')number_of_samples = len(_x)x_pos = np.array(_x)x_pos = x_pos.reshape(number_of_samples, 1)y_pos = np.array(_y)y_pos = y_pos.reshape(number_of_samples, 1)
# *****************************plt.scatter( x_pos, y_pos, color="gold", marker=".",label="Training data with noise and outliers")
plt.xlabel('x', fontsize=15)plt.ylabel('y', fontsize=15)
# ***************************
t1 = x_pos ** 2t1 = t1.reshape(number_of_samples, 1)t2 = x_pos.reshape(number_of_samples, 1)K = np.concatenate( (t1, t2), axis=1 )
# ************************from sklearn.linear_model import LinearRegressionreg = LinearRegression().fit(K, y_pos)
print( f'Linear regression model: Coefficients = {reg.coef_} intercept = {reg.intercept_}' )print( f' y = {reg.coef_[0][0]} * x**2 + {reg.coef_[0][1]} * x +{reg.intercept_}' )
# ****************************#Draw the fitted curvemin_x = min(x_pos)max_x = max(x_pos)
numOfPoint = 300temp_x = np.linspace(min_x, max_x, num = numOfPoint)xpoints = temp_x.reshape(numOfPoint, 1)xxpoints = xpoints ** 2H = np.concatenate( (xxpoints, xpoints), axis=1)prediction = reg.predict(H)
plt.plot(xpoints, prediction, color="blue", linewidth=1,label="LR model")
plt.legend(loc='lower left', fontsize=12)plt.title('Linear Regression - data with noise and outliers')plt.show()# ******************************prediction = reg.predict( K )residual = prediction - y_posplt.hist(residual, bins=30)plt.xlabel('Residual', fontsize=15)plt.ylabel('Count', fontsize=15)plt.title('Residual histogram (data with noise andoutliers)')plt.show()
# ****************************# Predict y value given x value of 40.0
x_test = np.array([[40.0 ** 2, 40.0],])y_result = reg.predict(x_test)
print(f'When x = {x_test[0]}, y = {y_result[0]}')
python python
"ython 3.9.12 (main, Apr 52022,01:53:17) ype "copyright", "credits" or "license" for more information. Python 8.2.0 - An enhanced Interactive Python. in [1]: runfile('/Users/andysin/Desktop/Big data EA/LinearRegression_ParabolicTest01.py', wdir= '/Users/ indysin/Desktop/Big data EA') inear regression model: Coefficients =[[−0.013650851.04143505]] intercept =[4.30564464] y=−0.0136508469172617∗x∗2+1.0414350542166018∗x+[4.30564464] Warning Figures now render in the Plots pane by default. To make them also appear inline in the Console, uncheck "Mute Inline Plotting" under the Plots pane options menu. When x=[1600.40.],y=[24.12169175] in [2]: