Adding line to scatter plot using python's matplotlib

Question

I am using python's matplotlib and want to create a matplotlib.scatter() with additional line. The line should proceed from the lower left corner to the upper right corner independent of the scatters content. A linear regression through the data, like in this post, is not what I am looking for. Also it should be dynamically and independent of the scatter input.

This should be the final plot:

EDIT:

Doing this got me the result:

# Scatter Plot
x = data_calc_hourly.temp
y =  data_obs_hourly.temp

lineStart = data_calc_hourly.temp.min() 
lineEnd = data_calc_hourly.temp.max()  

plt.figure()
plt.scatter(x, y, color = 'k', alpha=0.5)
plt.plot([lineStart, lineEnd], [lineStart, lineEnd], 'k-', color = 'r')
plt.xlim(lineStart, lineEnd)
plt.ylim(lineStart, lineEnd)
plt.show()

Is there any better way ?

You could get axes limits using get_ylim() and get_xlim() and then calculate the formula for the linear function you are looking for. — Luka, Nov 09 '16 at 22:01
Linear function is a straight independent line. This way it depends only on the axes matplotlib chooses. — Luka, Nov 09 '16 at 22:05

unutbu · Accepted Answer · 2016-11-09T22:20:43.733

33

This draws a diagonal line which is independent of the scatter plot data and which stays rooted to the axes even if you resize the window:

import numpy as np
import matplotlib.pyplot as plt
import matplotlib.lines as mlines
import matplotlib.transforms as mtransforms

x, y = np.random.random((2, 100))*2
fig, ax = plt.subplots()
ax.scatter(x, y, c='black')
line = mlines.Line2D([0, 1], [0, 1], color='red')
transform = ax.transAxes
line.set_transform(transform)
ax.add_line(line)
plt.show()

edited Nov 09 '16 at 22:20

answered Nov 09 '16 at 22:15

unutbu

842,883
184
1,785
1,677

exactly what I was looking for ! – Manuel Nov 10 '16 at 21:49
13

this can be done without the additional mlines import, just using the plot interface: ``ax.plot([0, 1], [0, 1], color='red', transform=ax.transAxes)`` – georgedeath Jul 29 '20 at 11:05

score 4 · Answer 2 · answered Mar 01 '19 at 10:05

Besides unutbu's answer one other option is to get the limits of the axis after you ploted the data and to use them to add the line. After this you will still need to change back the axis limits as they would change with the addition of the line:

# Scatter Plot
x = data_calc_hourly.temp
y =  data_obs_hourly.temp

lineStart = data_calc_hourly.temp.min() 
lineEnd = data_calc_hourly.temp.max()  

plt.figure()
plt.scatter(x, y, color = 'k', alpha=0.5)
y_lim = plt.ylim()
x_lim = plt.xlim()
plt.plot(x_lim, y_lim, 'k-', color = 'r')
plt.ylim(y_lim)
plt.xlim(x_lim)
plt.show()

score 0 · Answer 3 · answered Jul 18 '21 at 14:56

I have tried updating the min and max limits for the cases where X and Y axis have different max and min data.

x = data_calc_hourly.temp
y =  data_obs_hourly.temp

calc_min = data_calc_hourly.temp.min()
calc_max = data_calc_hourly.temp.max()

obs_min = data_obs_hourly.temp.min()
obs_max = data_obs_hourly.temp.max()

lineStart = min(calc_min,obs_min)
lineEnd = max(calc_max,obs_max)

plt.figure()
plt.scatter(x, y, color = 'k', alpha=0.5)
plt.plot([lineStart, lineEnd], [lineStart, lineEnd], color = 'r')
plt.xlim(lineStart, lineEnd)
plt.ylim(lineStart, lineEnd)
plt.show()

Adding line to scatter plot using python's matplotlib

3 Answers3

Linked