python-contour draws contour lines (contour lines)

1.Detailed explanation of matplotlib.pyplot.contour parameters

matplotlib.pyplot.contour([X, Y,] Z, [levels], **kwargs)

The
key parameters of Parameters are as follows, the remaining parameters can be found on the official website

X, Y: array-like, optional
Z coordinates of the median value.
X and Y must both be two-dimensional and have the same shape as Z (for example, created by numpy.meshgrid), or they must both be one-dimensional, so len(X) = M is the number of columns in Z, len(Y ) = N is the number of rows in Z.
If they are not given, they are assumed to be integer indexes, ie X = range(M), Y = range(N).

Z: array-like (N, M)
draws the contour height value

levels: int or array-like, optional
Determine the number and position of contour lines/regions

colors: color string or sequence of colors, optional
suitable for contour lines and contour areas

linestyles: {None,'solid','dashed','dashdot','dotted'}, optional
is only applicable to outlines. The line
style can be an iterable object that specifies a set of string styles to be used. If this iterable object is smaller than the number of contour layers, it will be repeated if necessary.

linewidths: float or array-like, default: rcParams[“contour.linewidth”] (default: None)
is only applicable to
the line width of contour lines.
If it is a number, all contour lines will be drawn with this line width.
If it is a sequence, the levels are drawn in ascending order, and the line widths are drawn in the specified order.
If not, return to rcParams["lines.linewidth"]

2. Examples

import warnings
import numpy as np
import pandas as pd
import matplotlib as mpl
from matplotlib import colors
import matplotlib.pyplot as plt
from sklearn.svm import SVC
from sklearn.metrics import accuracy_score


# 加载数据
def loaddata():
    data = pd.read_csv('data/svm3.txt', header=0, delimiter='\t')
    X = data.iloc[:, :2]
    y = data.iloc[:, 2]
    return X, y


if __name__ == '__main__':
    # 消除警告
    warnings.filterwarnings(action='ignore')
    # 设置样本显示格式
    np.set_printoptions(suppress=True)
    x, y = loaddata()
    # 分类器
    # 超参数为C、gamma
    clf_param = (('rbf', 1, 0.1), ('rbf', 1, 1), ('rbf', 1, 10), ('rbf', 1, 100),
                 ('rbf', 5, 0.1), ('rbf', 5, 1), ('rbf', 5, 10), ('rbf', 5, 100),
                 ('rbf', 1, 5), ('rbf', 50, 5), ('rbf', 100, 5), ('rbf', 1000, 5))
    x1_min, x2_min = np.min(x, axis=0)
    x1_max, x2_max = np.max(x, axis=0)
    x1, x2 = np.mgrid[x1_min:x1_max:200j, x2_min:x2_max:200j]
    grid_test = np.stack((x1.flat, x2.flat), axis=1)

    cm_light = mpl.colors.ListedColormap(['#77E0A0', '#FFA0A0'])
    cm_dark = mpl.colors.ListedColormap(['g', 'r'])
    mpl.rcParams['font.sans-serif'] = [u'SimHei']
    mpl.rcParams['axes.unicode_minus'] = False
    plt.figure(figsize=(14, 10), facecolor='w')
    for i, param in enumerate(clf_param):
        clf = SVC(C=param[1], kernel=param[0])
        clf.gamma = param[2]
        # if param[0] == 'rbf':
        #     clf.gamma = param[2]
        #     title = u'高斯核，C=%.1f，$\gamma$ =%.1f' % (param[1], param[2])
        # else:
        #     title = u'线性核，C=%.1f' % param[1]

        clf.fit(x, y)
        y_hat = clf.predict(x)
        print(u'准确率：', accuracy_score(y, y_hat))
        title = u'C=%.1f，gamma =%.1f,准确率1=%.2f' % (param[1], param[2], accuracy_score(y, y_hat))

        print(title)
        print(u'支撑向量的数目：', clf.n_support_)
        print(u'支撑向量的系数：', clf.dual_coef_)
        print(u'支撑向量：', clf.support_)

        # 画图
        plt.subplot(3, 4, i + 1)
        grid_hat = clf.predict(grid_test)  # 预测分类值
        grid_hat = grid_hat.reshape(x1.shape)  # 使之与输入的形状相同
        # 伪彩图
        plt.pcolormesh(x1, x2, grid_hat, cmap=cm_light, alpha=0.8)  # 画决策边界
        plt.scatter(x.iloc[:, 0], x.iloc[:, 1], c=y, edgecolors='k', s=40, cmap=cm_dark)  # 样本的显示
        plt.scatter(x.iloc[clf.support_, 0], x.iloc[clf.support_, 1], edgecolors='k', facecolors='none', s=100,
                    marker='o')  # 支撑向量

        # clf.decision_function与参数decision_function_shape取’ovr’、’ovo’有关，是点到超平面的函数间隔。程序首先是计算出’ovo’结果，然后聚合结果。
        z = clf.decision_function(grid_test)
        z = z.reshape(x1.shape)
        # contour绘制等高线图
        plt.contour(x1, x2, z, colors=list('kbrbk'), linestyles=['--', '--', '-', '--', '--'],
                    linewidths=[1, 0.5, 1.5, 0.5, 1], levels=[-1, -0.5, 0, 0.5, 1])

        plt.xlim(x1_min, x1_max)
        plt.ylim(x2_min, x2_max)
        plt.title(title, fontsize=14)

    plt.suptitle(u'SVM不同参数的分类', fontsize=20)
    # tight_layout会自动调整子图参数，使之填充整个图像区域
    plt.tight_layout(1.4)
    # 调整子图间距离
    plt.subplots_adjust(top=0.92)
    plt.show()