Python method for completing logistic regression

This article mainly describes the Python implementation of the method of logistic regression example, this is a machine learning course of an experiment, organized to share to everyone, the need for friends can refer to the study, down to see it together.

The principle of this paper is very simple, the optimization method is to use gradient descent. There are test results later.

Let's take a look at the example code for the implementation:

# coding=utf-8from Math Import expimport matplotlib.pyplot as Pltimport numpy as Npfrom Sklearn.datasets.samples_generato R Import make_blobsdef sigmoid (num): "':p Aram Num: The value after the x:return:sigmoid to be calculated" if type (num) = = int or type (num) = = f Loat:return 1.0/(1 + exp ( -1 * num)) else:raise valueerror, ' only int or float data can compute sigmoid ' class Logisti  C (): Def init (self, x, y): If type (x) = = Type (y) = = list:self.x = Np.array (x) self.y = Np.array (y) elif type (x) = =   Type (y) = = np.ndarray:self.x = x self.y = y else:raise valueerror, ' Input data error ' Def sigmoid (self, x): ' :p Aram x: Input vector: return: The vector result of simgoid computed for the input vector as a whole "s = Np.frompyfunc (lambda x:sigmoid (x), 1, 1) return S (x) def t Rain_with_punish (self, alpha, errors, punish=0.0001): "':p Aram Alpha:alpha for learning rate:p Aram errors: The threshold for stopping iterations when the error is less:p ARA M punish: penalty factor:p Aram times: Maximum iterations: return: "' self.punish = punish dimension = self.x.shape[1] Self.theta = Np.ra Ndom.random (Dimension) cOmpute_error = 100000000 times = 0 while compute_error > errors:res = Np.dot (self.x, self.theta) delta = self.si Gmoid (RES)-SELF.Y Self.theta = Self.theta-alpha * NP.DOT (self.x.t, Delta)-punish * Self.theta # with penalty gradient descent method Compu  Te_error = np.sum (delta) times + = 1 def predict (self, x): "':p Aram x: Gives a new unlabeled vector: return: Returns the category ' ' X ' as determined by the calculated parameter  Np.array (x) if Self.sigmoid (Np.dot (x, Self.theta)) > 0.5:return 1 else:return 0def test1 (): "For testing and drawing, showing effect : return: ' x, y = make_blobs (n_samples=200, centers=2, n_features=2, Random_state=0, center_box= (ten)) x1 = [] y1 = [] x2 = [] y2 = [] for i in range (len (y)): if y[i] = = 0:x1.append (x[i][0]) y1.append (x[i][1]) elif y[i] = = 1:x2. Append (X[i][0]) y2.append (x[i][1]) # above are processed data, generating two classes of data P = Logistic (x, y) p.train_with_punish (alpha=0.00001, errors=0.00 5, punish=0.01) # step is 0.00001, maximum allowable error is 0.005, penalty factor is 0.01 x_test = Np.arange (ten, 0.01) Y_test = ( -1 * p.theta[0]/p.theta[1 ]) * X_test Plt.plot (x_teSt, Y_test, c= ' G ', label= ' Logistic_line ') plt.scatter (x1, y1, c= ' R ', label= ' positive ') plt.scatter (x2, y2, c= ' b ', label= ' Negative ') Plt.legend (loc=2) plt.title (' punish value = ' + p.punish.str ()) plt.show () if name = = ' main ': Test1 ()

Running results such as

Summarize

"Recommended"

1. Python Free video tutorial

2. Python Basics Getting Started tutorial

3. Python Object-oriented video tutorial