Python: Barplot with colorbar

Question

I'm trying to make a bar plot with a color bar, each bar's hight is one variable (y) and each bar should have a color depending on another variable (c).

What I've got to is this (simple example):

data_x = [0,1,2,3]
data_hight = [60,60,80,100]
data_color = [1000,500,1000,900]


data_color = [x / max(data_color) for x in data_color]
fig, ax = plt.subplots(figsize=(15, 4))

my_cmap = plt.cm.get_cmap('GnBu')
colors = my_cmap(data_color)
rects = ax.bar(data_x, data_hight, color=colors)

CS = plt.contourf([data_x, data_color],cmap=my_cmap)

cbar = plt.colorbar(CS, cmap=my_cmap)
cbar.set_label('Color', rotation=270,labelpad=25)

plt.xticks(data_x)    
plt.ylabel("Y")

plt.show()

The main problem is that the histogram colors are fine but the color bar is in a diferent scale. besides that I can see a blue line at y=0, it shouldn't be there.

Any help will be a preciated. Thanks!

Maybe you could include an image illustrating what you mean? Also it remains unclear what you display and autolabel functions do. You should really try here to come up with a minimal example, that is stand-alone and focussed to the programming problem, not to your larger set of functions. — Tom de Geus
– Tom de Geus, Commented Jul 6, 2018 at 7:00

ImportanceOfBeingErnest · Accepted Answer · 2018-07-06 08:05:44Z

28

You are creating a contourf plot inside your bar plot. That makes no sense.

Instead you would need to create a mappable without any visual representation to supply to the colorbar. This would be a ScalarMappable.

import matplotlib.pyplot as plt
from matplotlib.cm import ScalarMappable

data_x = [0,1,2,3]
data_hight = [60,60,80,100]
data_color = [1000.,500.,1000.,900.]


data_color = [x / max(data_color) for x in data_color]
fig, ax = plt.subplots(figsize=(15, 4))

my_cmap = plt.cm.get_cmap('GnBu')
colors = my_cmap(data_color)
rects = ax.bar(data_x, data_hight, color=colors)

sm = ScalarMappable(cmap=my_cmap, norm=plt.Normalize(0,max(data_color)))
sm.set_array([])

cbar = plt.colorbar(sm)
cbar.set_label('Color', rotation=270,labelpad=25)

plt.xticks(data_x)    
plt.ylabel("Y")

plt.show()

answered Jul 6, 2018 at 8:05

ImportanceOfBeingErnest

342k61 gold badges737 silver badges771 bronze badges

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

3 Comments

Andy Over a year ago

How do you include the colorbar as a legend (i.e horizontal on the top right of the figure)?

Toonia Over a year ago

How can you do this from a dataframe, the case of multiple bar charts?

Patafikss Over a year ago

ValueError: Unable to determine Axes to steal space for Colorbar

WhatTheDuck · Accepted Answer · 2024-09-27 17:06:13Z

For matplotlib versions greater than ~3.6, to avoid the "ValueError: Unable to determine Axes to steal space for Colorbar" error, you need to specify the axes for the colorbar, which can be done via plt.gca():

import matplotlib.pyplot as plt
from matplotlib.cm import ScalarMappable

data_x = [0,1,2,3]
data_hight = [60,60,80,100]
data_color = [1000.,500.,1000.,900.]


data_color = [x / max(data_color) for x in data_color]
fig, ax = plt.subplots(figsize=(15, 4))

my_cmap = plt.get_cmap('GnBu') # plt.cm.get_cmap is deprecated as of v3.7
colors = my_cmap(data_color)
rects = ax.bar(data_x, data_hight, color=colors)

sm = ScalarMappable(cmap=my_cmap, norm=plt.Normalize(0,max(data_color)))
sm.set_array([])

cbar = plt.colorbar(sm, ax = plt.gca()) # Supply the current axes to the colorbar
cbar.set_label('Color', rotation=270,labelpad=25)

plt.xticks(data_x)    
plt.ylabel("Y")

plt.show()

Collectives™ on Stack Overflow

Python: Barplot with colorbar

2 Answers 2

3 Comments

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

3 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related