Change colour of colorbar in Python Matplotlib

Question

I have a code that gives me a scatter plot of predicted vs actual values as a function of concentration. The data is pulled from an excel csv spreadsheet.

This is the code:

import matplotlib.pyplot as plt
from numpy import loadtxt


dataset = loadtxt("ColorPlot.csv", delimiter=',')
x = dataset[:,0]
y = dataset[:,1]
z = dataset[:,2]

scaled_z = (z - z.min()) / z.ptp()

colors = plt.cm.viridis(scaled_z)
sc=plt.scatter(x, y, c=colors)
plt.clim(0, 100)
plt.colorbar()
plt.xlabel("Actual")
plt.ylabel("Predicted")

plt.show()

And with this I get a nice graph:

However if I change the color to something like

colors = plt.cm.plasma(scaled_z)

I get the graph below but the colorbar remains unchanged.

I've tried lots of different things like cmap or edgecolors but I don't know how to change it. And I want to keep the code as simple as it currently is because I want to readily change the third variable of z based on my excel spreadsheet data.

Is there also a way for the scale of the colorbar to pick up what the scale is from the excel spreadsheet without me manually specifying 0-100?

The standard way is to call it as plt.scatter(x, y, c=z, cmap='plasma'). And leaving out plt.clim(). In that standard way, the colorbar will get the 'plasma' colors, and will have indications of the lowest and highest z values. If, instead, you want the colorbar to go between 0 and 100, you could do scaled_z = 100*(z - z.min()) / z.ptp() and plt.scatter(x, y, c=scaled_z, cmap='plasma'). In any case, it is best to leave out plt.clim() — JohanC
– JohanC, Commented Sep 21, 2020 at 16:11
Omg thats perfect. Literally solved it straight away. Thanks so much. I removed the scaled z term, put cmap in the scatter term and removed plt.clim(). Perfect thanks @JohanC — paula.n
– paula.n, Commented Sep 21, 2020 at 16:40

Trenton McKinney · Accepted Answer · 2021-10-04 14:17:51Z

12

To get the right color bar, use the following code:

colormap = plt.cm.get_cmap('plasma') # 'plasma' or 'viridis'
colors = colormap(scaled_z)
sc = plt.scatter(x, y, c=colors)
sm = plt.cm.ScalarMappable(cmap=colormap)
sm.set_clim(vmin=0, vmax=100)
plt.colorbar(sm)
plt.xlabel("Actual")
plt.ylabel("Predicted")
plt.show()

For my random generated data I got the following plot:

Now replace 'plasma' with 'viridis' and check the other variant.

edited Oct 4, 2021 at 14:17

Trenton McKinney

63.2k41 gold badges169 silver badges212 bronze badges

answered Sep 21, 2020 at 16:45

Valdi_Bo

31.1k4 gold badges29 silver badges45 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

Jody Klymak · Accepted Answer · 2020-09-21 17:44:36Z

1

You should not scale your data, unless you want the colorbar to be incorrect. Once you have the PathCollection from the scatter call, you can call set_cmap and set_clim on that and the colorbar should track. (you could also explicitly associate the colorbar with the PathCollection to avoid ambiguity)

import matplotlib.pyplot as plt
import numpy as np


x = np.random.randn(100)
y = np.random.randn(100)
z = np.random.randn(100)

sc=plt.scatter(x, y, c=z, cmap='viridis')
plt.clim(0, 100)
plt.colorbar(sc)
plt.xlabel("Actual")
plt.ylabel("Predicted")
sc.set_cmap('plasma')
sc.set_clim(-1, 1)

plt.show()

answered Sep 21, 2020 at 17:44

Jody Klymak

6,1572 gold badges20 silver badges37 bronze badges

Comments

Liris · Accepted Answer · 2020-09-21 16:10:48Z

0

Your code return for me an error TypeError: You must first set_array for mappable ...

The following is a simplest syntax that works for me:

import matplotlib.pyplot as plt
import numpy as np

a = np.random.random(100)
b = np.random.random(100)

scaled_z = (a + b)/a

plt.figure()
plt.scatter(a, b, c = scaled_z, cmap = 'plasma') ## you can directly change the colormap here
plt.colorbar()
plt.tight_layout()
plt.show()

answered Sep 21, 2020 at 16:10

Liris

1,5463 gold badges16 silver badges35 bronze badges

Comments

Friedrich -- Слава Україні · Accepted Answer · 2025-05-03 09:55:02Z

0

Simply use the cmap keyword.

x = np.random.randn(100)
y = np.random.randn(100)
c = np.random.randn(100)

plt.scatter(x, y, c=c, cmap='plasma')
plt.colorbar()

Unfortunately, edgecolors would need more work.

answered May 3 at 9:55

Friedrich -- Слава Україні

3,9831 gold badge34 silver badges53 bronze badges

Collectives™ on Stack Overflow

Change colour of colorbar in Python Matplotlib

4 Answers 4

Comments

Comments

Comments

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

4 Answers 4

Comments

Comments

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related