多项式回归绘图问题

Question

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from sklearn.preprocessing import PolynomialFeatures
from sklearn.metrics import mean_squared_error, r2_score
import seaborn as sns
from sklearn import metrics
from matplotlib.pylab import rcParams
from sklearn.preprocessing import MinMaxScaler
from sklearn.linear_model import LinearRegression

df = pd.read_csv('FB.csv')
datapol = pd.DataFrame(df, columns=['Date', 'Close'])

ptrain, ptest = train_test_split(datapol, test_size=0.20, random_state=42)
px_train = np.array(ptrain.index).reshape(-1, 1)
py_train = ptrain['Close']

poly= PolynomialFeatures(degree=4)
X_train_poly= poly.fit_transform(px_train)
pol_train_reg = LinearRegression()
pol_train_reg.fit(X_train_poly, py_train)
py_train_pred = pol_train_reg.predict(X_train_poly)

plt.figure(1, figsize=(16, 10))
plt.title('Linear Regression | Price vs Time')
plt.scatter(px_train, py_train, edgecolor='w', label='Actual Price')
plt.plot(px_train, pol_train_reg.predict(poly.fit_transform(px_train)), color='r', label='Polynomial')
plt.xlabel('Integer Date')
plt.ylabel('Stock Price')
plt.legend()
plt.show()`

This is the method of polynomial regression .The problem is plotting the graph. Here is the graph. Can anyone tell me how can i fix this ?

https://i.stack.imgur.com/gefcY.png

Answer 1

I don't have access to your dataset so I had to create one from another tutorial Basically I think you need to sort your values otherwise it creates the "block" you've seen.

So, foremost, I had to add these lines:

sort_axis = operator.itemgetter(0)
sorted_zip = sorted(zip(x,y_poly_pred), key=sort_axis)
x, y_poly_pred = zip(*sorted_zip)

Full code:

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from sklearn.preprocessing import PolynomialFeatures
from sklearn.metrics import mean_squared_error, r2_score
import seaborn as sns
from sklearn import metrics
from matplotlib.pylab import rcParams
from sklearn.preprocessing import MinMaxScaler
from sklearn.linear_model import LinearRegression

np.random.seed(0)
x = 2 - 3 * np.random.normal(0, 1, 20)
y = x - 2 * (x ** 2) + 0.5 * (x ** 3) + np.random.normal(-3, 3, 20)

# transforming the data to include another axis
px_train = x[:, np.newaxis]
y = y[:, np.newaxis]

poly= PolynomialFeatures(degree=4)
x_train_poly = polynomial_features.fit_transform(px_train)

X_train_poly= poly.fit_transform(px_train)
pol_train_reg = LinearRegression()
pol_train_reg.fit(X_train_poly, y)
py_train_pred = pol_train_reg.predict(X_train_poly)

plt.figure(1, figsize=(16, 10))
plt.title('Linear Regression | Price vs Time')
plt.scatter(px_train, y, edgecolor='w', label='Actual Price')

sort_axis = operator.itemgetter(0)
sorted_zip = sorted(zip(x,y_poly_pred), key=sort_axis)
x, y_poly_pred = zip(*sorted_zip)

plt.plot(x, y_poly_pred, color='m')
plt.xlabel('Integer Date')
plt.ylabel('Stock Price')
plt.legend()
plt.show()

This will generate a proper graph. Admittedly the regression needs tuning, but that was not the point of the question: