python绘制3D瀑布图

成品：

代码：

import matplotlib.pyplot as plt
import matplotlib.ticker as ticker
from mpl_toolkits.mplot3d.art3d import Poly3DCollection
import numpy as npdef line_3d(x, y, z, x_label_indexs):"""在y轴的每个点，向x轴的方向延伸出一个折线面：展示每个变量的时序变化。x: x轴，时间维，右边。y: y轴，变量维，左边。z: z轴，数值维。二维矩阵，y列x行。每一行是对应变量的一个时间序列。x_label_indexs: 需要标注的时间点。"""x_num = len(x)y_num = len(y)if z.shape[0] != y_num or z.shape[1] != x_num:return -1# 制作坐标格点(z中每个点对应的x、y坐标)X, Y = np.meshgrid(x, y)# 初始化canvas = plt.figure()  # 创建画布axs = canvas.add_subplot(111, projection='3d')  # 添加三维子图# 若把111改成234，则意思是：创建一个2*3的网格，并在第4个格子中创建一个axes# 绘制折线面for i in range(y_num):  # 遍历# z值线，即实际数据。axs.plot(Y[i], X[i], z[i], color=plt.cm.viridis(i/y_num),linestyle='-', linewidth=1, marker='o', markersize=3, alpha=0.3)# 0值线（z=0），与“地面”连接。axs.plot(Y[i], X[i], np.zeros_like(z[i]), color='gray', alpha=0.5)# 绘制有颜色的平面：本质是填充z值与0值之间的区域。polygon = [[Y[i, 0], X[i, 0], 0],    # 左下[Y[i, -1], X[i, -1], 0],  # 右下]for j in range(x_num-1, -1, -1):  # 依次添加点，使得polygon成为一个完整的闭合多边形polygon.append([Y[i, j], X[i, j], z[i, j]])axs.add_collection3d(Poly3DCollection([polygon], color=plt.cm.viridis(i/y_num), alpha=0.3))# 标注数字（z值）for k in x_label_indexs:axs.text(Y[i, k]-0.05, X[i, k], z[i, k]+0.02, f'{z[i, k]:.2f}',color='black', ha='center', size=7)# 用虚线将需要标注的时间（y）连起来for k in x_label_indexs:axs.plot(Y[:, k], X[:, k], z[:, k], linestyle='--', linewidth=0.8, color='gray')axs.grid()plt.show()if __name__ == '__main__':x = np.arange(5)time = np.arange(1, 15, 2)z = np.array([[0.20, 0.34, 0.38, 0.43, 0.44, 0.50, 0.61],[0.21, 0.40, 0.38, 0.43, 0.60, 0.72, 0.75],[0.22, 0.43, 0.44, 0.60, 0.77, 0.84, 0.92],[0.23, 0.42, 0.44, 0.43, 0.64, 0.77, 0.86],[0.38, 0.42, 0.43, 0.49, 0.55, 0.60, 0.81]])line_3d(time, x, z, [1, 4, 6])