Numpy XYZ Scatter 3d autoscale_xyz

Community
1

Hi,
I have a numpy matrix (100 000 x 3) each lines correspond to coordonnates XYZ of one point. When I try to plot it in 3D, I am using Axe3d,. Scatter. After that I have flatten my matrix just before using autoscale_xyz. And show at the end. This graph looks good.

But when I apply minmaxscaller on my XYZ matrix before plot, the autoscale_xyz doesn’t work and the graph doesn’t look good.

Can you help me?

2

Hi, can you please provide some small sample code that reproduces the issue?

3

Hi,

Thanks for your help.
Of course :

from mpl_toolkits.mplot3d import Axes3D
from sklearn.preprocessing import MinMaxScaler, StandardScaler, RobustScaler

fig = plt.figure(figsize=(12, 8))
ax = Axes3D(fig)
ax.scatter(XYZ[:, 0], XYZ[:, 1], XYZ[:,2])
scale = XYZ.flatten() # Auto scale to the mesh size
ax.auto_scale_xyz(scale, scale, scale)
fig.show()

When I do the transformation below and try to plot it with the same script than above, the plot doesn’t make sens.

XYZ_scale = MinMaxScaler().fit_transform(XYZ)