How to show a confidence interval in python using matplotlib?

Question

I need to show a confidence interval, like in this image:

but I don't know how to do it. I've tried doing lb.fill_between(x, (y1-ci), (y1+ci), color = 'b', alpha = 0.1) but it returns the error: AttributeError: 'list' object has no attribute 'fill_between'.
This is my code:

import matplotlib.pyplot as plt


x = [10, 100, 1000]
y1 = [215103, 22824279.7, 22063128311]
y2 = [211298.5, 21315505.2, 20563930722]

plt.subplot(2, 2, 1)

ci = 1300
#la = plt.plot(x,y,'b*', label = 'normal')
lb = plt.plot(x,y1, '#FA8072', label = 'LI')
lc = plt.plot(x,y2, '#7FFFD4', label = 'LU')
plt.legend(loc = 'upper left')
plt.title("L1-dcache-loads")
plt.xscale("log")

Thanks in advance!

lb is a list. You need to plt.fill_between. Note that y1-ci will also give you an error — DavidG
– DavidG, Commented Apr 20, 2020 at 14:43
@DavidG when i do plt.fill_between I get this error: unsupported operand type(s) for -: 'list' and 'int', just like you said. How can I fix this? — Ana Laura Chioca Vieira
– Ana Laura Chioca Vieira, Commented Apr 20, 2020 at 14:47

gnodab · Accepted Answer · 2020-04-20 14:55:54Z

1

Use fill_between like this:

plt.figure()
plt.fill_between(x, y1, y2, edgecolor='g', facecolor='g', alpha=0.3)

y1 is your lower bound curve, and y2 is your upper bound curve. Output:

In your example:

plt.figure()
plt.plot(x,y1, '#FA8072', label = 'LI')
plt.fill_between(x, np.array(y1)-ci, np.array(y1)+ci, edgecolor='r', facecolor='r', alpha=0.3)
plt.plot(x,y2, '#7FFFD4', label = 'LU')
plt.fill_between(x, np.array(y2)-ci, np.array(y2)+ci, edgecolor='g', facecolor='g', alpha=0.3)
plt.legend(loc = 'upper left')
plt.title("L1-dcache-loads")
plt.xscale("log")

But the intervals are too small to see.

edited Apr 20, 2020 at 14:55

answered Apr 20, 2020 at 14:47

gnodab

8787 silver badges16 bronze badges

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

Collectives™ on Stack Overflow

How to show a confidence interval in python using matplotlib?

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related