Représentation gaussiennes en 3D

Le but de ce bout de code est de générer 3 gaussiennes et d'en faire la représentation graphique en 3D.

CODE PYTHON

step = 200

xx = np.linspace(0, 4, step) yy = xx Xg, Yg = np.meshgrid(xx, yy)

rho = cov[1, 0] / np.sqrt(cov[1,1] * cov[0,0]) rho_large = cov_large[1, 0] / np.sqrt(cov_large[1, 1] * cov_large[0, 0])

Z1 = plt.mlab.bivariate_normal(Xg, Yg, sigmax=noise_level, sigmay=noise_level, mux=mean_0[0], muy=mean_0[1], sigmaxy=rho * noise_level * noise_level)

Z2 = plt.mlab.bivariate_normal(Xg, Yg, sigmax=noise_level_large, sigmay=noise_level_large, mux=mean_1[0], muy=mean_1[1], sigmaxy=rho_large * noise_level_large * noise_level_large)

Z3 = plt.mlab.bivariate_normal(Xg, Yg, sigmax=noise_level, sigmay=noise_level, mux=mean_2[0], muy=mean_2[1], sigmaxy=rho * noise_level * noise_level)

fig3 = plt.figure(figsize=(9, 6), dpi = 90)

ax = fig3.add_subplot(111, projection='3d')

ax.plot_surface(Xg, Yg, (Z1 + Z2 + Z3) / 3, cmap='Oranges', rstride=3, cstride=3, alpha=0.9, linewidth=0.5)