Specifying matplotlib colors for a call to axis.plot() with multiple lines

Question

I would like to plot a 3xN matrix to obtain three lines, specifying the colors for each one. I can do this with three separate calls to plot() :

fig = plt.figure()
theta = np.arange(0,1,1.0/720) * 2 * np.pi
abc = np.cos(np.vstack([theta, theta+2*np.pi/3, theta+4*np.pi/3])).T
ax = fig.add_subplot(1,1,1)
ax.plot(theta,abc[:,0],color='red')
ax.plot(theta,abc[:,1],color='green')
ax.plot(theta,abc[:,2],color='blue')

Is there a way to do the same thing, specifying colors with one call to plot() ?

ax.plot(theta,abc,????)

I've tried

ax.plot(theta,abc,color=['red','green','blue'])

but I get ValueError: could not convert string to float: red in a callback.

Answer 1

As far as I know, that's not possible. According to the documentation the color= argument must be a single color (or an array of 3-4 floats). Is there a particular reason you need to do only one call?

If it must be a one-liner, you could do the following:

ax.plot(theta,abc[:,0],color='red',theta,abc[:,1],color='green',theta,abc[:,2],color='blue')

which is one line, but does not really save on typing.

Or:

for i,c in enumerate(['red','green','blue']):
    ax.plot(theta,abc[:,i],color=c)

EDIT

Actually, I thought of one way to do this. If you simply call plot(theta, abc[:,:]) , the colors will be cycled through the list of colors in the rcParams axes.prop_cycle .

Therefore, you could (temporarily) change the color cycle to red,green,blue before calling your plot command.

from cycler import cycler

#save the current color cycler
old = matplotlib.rcParams['axes.prop_cycle']

#set the color cycler to r,g,b
matplotlib.rcParams['axes.prop_cycle'] = cycler(color=['r','g','b'])

fig = plt.figure()
theta = np.arange(0,1,1.0/720) * 2 * np.pi
abc = np.cos(np.vstack([theta, theta+2*np.pi/3, theta+4*np.pi/3])).T
ax = fig.add_subplot(1,1,1)
ax.plot(theta,abc)  # <----- single call to `plot` for the 3 lines

#restore original color cycle
matplotlib.rcParams['axes.prop_cycle'] = old

Answer 2

An option is to change the color cycle from which colors are chosen automatically, such that it contains those colors needed for the plot.

ax.set_prop_cycle('color', ["red", "green","blue"]) 
ax.plot(theta,abc)

This can also be done using rcParams, see eg here .

Complete example:

import matplotlib.pyplot as plt
import numpy as np

fig = plt.figure()
theta = np.arange(0,1,1.0/720) * 2 * np.pi
abc = np.cos(np.vstack([theta, theta+2*np.pi/3, theta+4*np.pi/3])).T
ax = fig.add_subplot(1,1,1)

ax.set_prop_cycle('color', ["red", "green","blue"]) 
ax.plot(theta,abc)

plt.show()