How to plot multiple 2D Series in 3D (Waterfall plot) in Matplotlib

In this article, we will learn how to plot multiple 2D series data in a single 3D plot like the one shown below!!

Step 1: Importing required libraries (or only a few objects)

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

Step 2: Defining a figure and axis

fig = plt.figure()
ax = fig.add_subplot(projection=’3d’)

Step 3: Getting some data

# Plot a sin curve using the x and y axes.
for z in range(0, 50, 5):
    x = np.linspace(0, 1, 100)
    y = z*np.sin(x * 2 * np.pi) / 50 / 2 + 0.5
    plot2din3d(x, y, z)

Step 4: plot2din3d()

def plot2din3d(x,y,z):
    ax.plot(x, y, zs=z, zdir='z')
    2d_col_obj = ax.fill_between(x, 0.5, y, step='pre', alpha=0.1) 
    ax.add_collection3d(2d_col_obj, zs = z, zdir = 'z')

Step 5: Set Legend, set axes limits and labels

#ax.legend()
ax.set_xlim(0, 1)
ax.set_ylim(0, 1)
ax.set_zlim(0, 50)
ax.set_xlabel(‘X’)
ax.set_ylabel(‘Y’)
ax.set_zlabel(‘Z’)

Step 6: Customizing View Angle

ax.view_init(elev=20., azim=-35)

Step 7: Plot the plot

plt.show()

Final Result

We can do this for any type of series data!!

#The data above was, however, dense. Hence, looks continuous.

#How about discrete data, not so densely spaced?

Let us, set some values to some variables, which we can tweak to obtain different plots!!

num_samples = 10
num_series = 3
minx, maxx = 0, 10
res = 100
zmin, zmax = 0, 60
zinterval = 20

Next,

y = np.random.random((num_series,num_samples))

x = np.linspace(minx, maxx, num_samples)

z_idx_range = int((zmax-zmin)/zinterval)

colors = plt.cm.get_cmap(‘inferno’,z_idx_range)

Next, we interpolate points between the samples data points, to get a continuous looking plot!!

for i in range(z_idx_range):
    z = zinterval*i + zmin
    yz, ys, ye = y[i,:], None, None
    ynew_ = np.array([])
    for j in range(y.shape[1]-1):
        ys = y[i,j]
        ye = y[i,j+1]
        ynew_ = np.append(ynew_, np.linspace(ys, ye, res), axis = 0)

    xz = np.linspace(minx, maxx, (y.shape[1]-1)*res)
    plot2din3d_discrete(xz,x,ynew_,yz,z,colors(i))

plot2din3d_discrete()

def plot2din3d_discrete(x,xo,y,yo,z,c):
    ax.plot(xo, yo,’o’,color=c,zs=z, zdir=’z’)
    ax.plot(x, y, color=c,zs=z, zdir=’z’)
    # Add a 3D collection object to the plot.
    2d_col_obj = ax.fill_between(x, 0.0, y, step='mid', alpha=0.3)       
    ax.add_collection3d(2d_col_obj, zs = z)

Finishing Touches!!!

Steps 5 to 7 in the first example.

Result

References:

1. https://github.com/artmenlope/matplotlib-fill_between-in-3D/blob/master/FillBetween3d.py
2. https://matplotlib.org/2.0.2/mpl_toolkits/mplot3d/api.html
3. https://matplotlib.org/3.5.1/api/_as_gen/matplotlib.axes.Axes.fill_between.html
4. https://www.tutorialspoint.com/how-to-fill-the-area-under-a-step-curve-using-pyplot-matplotlib#:~:text=To%20fill%20the%20area%20under%20the%20curve%2C%20put%20x%20and,figure%2C%20use%20show()%20method.

Clap, Follow and Share if you found this post useful!!

Originally published on Medium.