Drawing a 3d scatter plot with data dependent markers and colors

Question

I have a dataframe df with 5 columns, f1,f2,f3,f4,y, where all values in all columns are from a finite set of integers, in fact, all columns are categorial columns, converted to integers. What I would like to do is to draw a 3d scatter plot, with f1,f2,f3 on the axis, the marker style should be determined by f4, and finally the color should be determined by the y column.

The following code deals with the axis and the colors.

import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
fig =plt.figure(figsize=(20,16)).gca(projection='3d')
fig.scatter(df['f1'], df['f2'], df['f3'], c=df['y'], s=100)
plt.show()

I'm however not sure how to get the marker style based on column f4. Remotely inspired by this post I would probably define a list of possible markers:

marker_styles = ['.','o','v','^','>','<','s','p','*','h','H','D','d','1']

Then I would group my data based on f4. For each group I would use the next marker, and rotate back to the beginning of the marker_styles list if there are more groups than markers.

I'm not sure how to execute this idea, or whether there are better alternatives.

Answer 1

You could iterate through all possible values for f4, create a filter for that value and combine it with a marker:

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

N = 100
print(np.random.randint(1, 10, N))
df = pd.DataFrame({'f1': np.random.randint(1, 11, N),
                   'f2': np.random.randint(1, 11, N),
                   'f3': np.random.randint(1, 11, N),
                   'f4': np.random.randint(1, 11, N),
                   'y': np.random.randint(1, 11, N)})
marker_styles = ['.', 'o', 'v', '^', '>', '<', 's', 'p', '*', 'h', 'H', 'D', 'd', '1']

fig = plt.figure(figsize=(20, 16)).gca(projection='3d')
f4min = df['f4'].min()
f4max = df['f4'].max()
for f, m in zip(range(f4min, f4max + 1), itertools.cycle(marker_styles)):
    filter = df['f4'] == f
    fig.scatter(df['f1'][filter], df['f2'][filter], df['f3'][filter], c=df['y'][filter], s=100, marker=m, cmap='plasma')
plt.show()