Linear regression example using matplotlib instead of pylab

Question

Since pylab is discouraged, I intend to use matplotlib from this example in https://www.wired.com/2011/01/linear-regression-with-pylab/

from pylab import *

x = [0.2, 1.3, 2.1, 2.9, 3.3]
y = [3.3, 3.9, 4.8, 5.5, 6.9]

(m, b) = polyfit(x, y, 1)
print(m, b)

yp = polyval([m, b], x)
plt.plot(x, yp)
plt.grid(True)
plt.scatter(x,y)
xlabel('x')
ylabel('y')
show()

If I start this one with

import matplotlib.pyplot as plt

I don't know how to replace polyfit and polyval functions in matplotlib. In line 4 and 7, these call the polyfit and polyval function (those are in the pylab module). What functions should I use instead for using matplotlib?

I want to use this example but using matplotlib.

Please don’t paste images of your code. Include the actual code so people trying to help you don’t have to transcribe your image. — Mark, Sep 07 '18 at 18:12
You are right. The site actually has the image, I'll try typing the code next time. — Rakibul Hassan, Sep 07 '18 at 18:17
FYI, you can edit questions and type the code yourself. @MarkMeyer was nice enough to use his own time to do it this time. — Hatt, Sep 07 '18 at 18:26
It looks good @RakibulHassan -- I didn't downvote this, but if I had I would undo it. — Mark, Sep 07 '18 at 19:08
@mark-meyer I don't mind about downvoting it. You responded properly what I really wanted. Much appreciated for what you have done. I didn't expect you would would type it for me. Thank you. — Rakibul Hassan, Sep 07 '18 at 19:17

score 3 · Accepted Answer · answered Sep 07 '18 at 18:25

3

polyfit and polyval are both available in numpy. So you can simple use:

import numpy as np
import matplotlib.pyplot as plt

x = [0.2, 1.3, 2.1, 2.9, 3.3]
y = [3.3, 3.9, 4.8, 5.5, 6.9]

(m, b) = np.polyfit(x, y, 1)
print(m, b)

yp = np.polyval([m, b], x)
plt.plot(x, yp)
plt.grid(True)
plt.scatter(x,y)

answered Sep 07 '18 at 18:25

Mark

90,562
7
108
148

Outstanding! As you said, polyfit and polyval are both available in numpy, we could go ahead and use it. Meanwhile, I was planning to use `import numpy as np import matplotlib.pyplot as plt x = [0.2, 1.3, 2.1, 2.9, 3.3] y = [3.3, 3.9, 4.8, 5.5, 6.9]` to plot and add the best fit line. What could be the best way to plot both with less code? Thank you once again @mark-meyer – Rakibul Hassan Sep 07 '18 at 18:37
Can this be done with Seaborn's regplot or lmplot as shown in https://stackoverflow.com/a/40478505 ? – Rakibul Hassan Sep 07 '18 at 18:53
I'm sorry for the delay. I thought I accepted both the answers. It seems that, you can only accept one. I didn't want to hurt anyone. – Rakibul Hassan Sep 07 '18 at 19:11

score 2 · Answer 2 · answered Sep 07 '18 at 18:27

2

The plotting functions are in matplotlib.pyplot but PyLab also includes numerical functions from NumPy which you can with

import numpy as np

np.polyfit(...)  # etc.

(See also the first answer here.)

answered Sep 07 '18 at 18:27

cheersmate

2,385
4
19
32

Linear regression example using matplotlib instead of pylab

2 Answers2