Pandas dataframe 3d visualization with different colors

user_17126106 0
Viewed 142

Imagine a dataframe like this:

i   X     Y     Z     label  
1   23    45    23    0
2   56    67    24    0
3   34    87    25    0
4   43    78    26    0
5   45    45    37    1
6   34    98    38    1
7   23    45    39    1
8   34    76    40    1
9   54    87    41    1

I know how to visualize x,y,z using matplot, but the problem is I want to set the color of each data using the label column, for example all 0 labeled rows should be in green and 1 labeled rows should be orange. I'm pretty new to python and it would be grate if give an example implementation.

Thank you so much for your help.

pandas matplotlib python
1 Answers

You can do this with set color for each label and plot scatter like below:

import matplotlib.pyplot as plt
import pandas as pd
import numpy as np
from mpl_toolkits.mplot3d import Axes3D

np.random.seed(365)
df = pd.DataFrame({
    'X': np.random.rand(200),
    'Y': np.random.rand(200),
    'Z': np.random.rand(200),
    'label' : np.hstack((np.zeros(100),np.ones(100)))})

fig = plt.figure(figsize=(12,7))
ax = fig.add_subplot(projection = '3d')

colors = {0 : 'orange', 1:'g'}
for l in df['label'].unique():
    ax.scatter(xs = df.loc[df.label == l, 'X'], 
               ys = df.loc[df.label == l, 'Y'],
               zs = df.loc[df.label == l, 'Z'], 
               color = colors[l])

plt.show()

Output:

enter image description here

user_17405770
Related