mirror of
https://github.com/Visualize-ML/Book4_Power-of-Matrix.git
synced 2026-02-03 02:24:03 +08:00
Add files via upload
This commit is contained in:
78
Book4_Ch19_Python_Codes/Bk4_Ch19_01.py
Normal file
78
Book4_Ch19_Python_Codes/Bk4_Ch19_01.py
Normal file
@@ -0,0 +1,78 @@
|
||||
|
||||
###############
|
||||
# Authored by Weisheng Jiang
|
||||
# Book 4 | From Basic Arithmetic to Machine Learning
|
||||
# Published and copyrighted by Tsinghua University Press
|
||||
# Beijing, China, 2022
|
||||
###############
|
||||
|
||||
|
||||
# Bk4_Ch19_01.py
|
||||
|
||||
import sympy
|
||||
|
||||
#define symbolic vars, function
|
||||
x1,x2 = sympy.symbols('x1 x2')
|
||||
# f_x = 3*x1**2-5*x2**2
|
||||
f_x = x1 - x2
|
||||
|
||||
#take the gradient symbolically
|
||||
grad_f = [sympy.diff(f_x,var) for var in (x1,x2)]
|
||||
|
||||
f_x_fcn = sympy.lambdify([x1,x2],f_x)
|
||||
|
||||
#turn into a bivariate lambda for numpy
|
||||
grad_fcn = sympy.lambdify([x1,x2],grad_f)
|
||||
|
||||
import numpy as np
|
||||
import matplotlib.pyplot as plt
|
||||
|
||||
xx1, xx2 = np.meshgrid(np.linspace(-4,4,41),np.linspace(-4,4,41))
|
||||
|
||||
# coarse mesh
|
||||
xx1_, xx2_ = np.meshgrid(np.linspace(-4,4,10),np.linspace(-4,4,10))
|
||||
V = grad_fcn(xx1_,xx2_)
|
||||
V_z = np.ones_like(V[1]);
|
||||
|
||||
ff_x = f_x_fcn(xx1,xx2)
|
||||
# ff_x_ = f_x_fcn(xx1_,xx2_)
|
||||
|
||||
color_array = np.sqrt(V[0]**2 + V[1]**2)
|
||||
l_3D_vectors = np.sqrt(V[0]**2 + V[1]**2 + V_z**2)
|
||||
|
||||
# 3D
|
||||
ax = plt.figure().add_subplot(projection='3d')
|
||||
ax.plot_wireframe(xx1, xx2, ff_x, rstride=1,
|
||||
cstride=1, color = [0.5,0.5,0.5],
|
||||
linewidth = 0.2)
|
||||
ax.contour3D(xx1, xx2, ff_x,20)
|
||||
ax.contour3D(xx1, xx2, ff_x, levels = 0, colors = 'k')
|
||||
|
||||
# plt.quiver (xx1_, xx2_, ff_x_, V[0], V[1], V_z,length = .5)
|
||||
ax.xaxis.set_ticks([])
|
||||
ax.yaxis.set_ticks([])
|
||||
ax.zaxis.set_ticks([])
|
||||
plt.xlim(-4,4)
|
||||
plt.ylim(-4,4)
|
||||
ax.view_init(30, -125)
|
||||
ax.set_xlabel('$x_1$')
|
||||
ax.set_ylabel('$x_2$')
|
||||
ax.set_zlabel('$f(x_1,x_2)$')
|
||||
plt.tight_layout()
|
||||
|
||||
|
||||
# 2D
|
||||
fig, ax = plt.subplots()
|
||||
plt.quiver (xx1_, xx2_, V[0], V[1], color_array,
|
||||
angles='xy', scale_units='xy',scale=2,
|
||||
edgecolor='none', facecolor= 'b')
|
||||
|
||||
plt.contour(xx1, xx2, ff_x,20)
|
||||
plt.contour(xx1, xx2, ff_x, levels = 0, colors = 'k')
|
||||
plt.show()
|
||||
ax.set_aspect('equal')
|
||||
ax.xaxis.set_ticks([])
|
||||
ax.yaxis.set_ticks([])
|
||||
ax.set_xlabel('$x_1$')
|
||||
ax.set_ylabel('$x_2$')
|
||||
plt.tight_layout()
|
||||
Reference in New Issue
Block a user