Add image to background of plot with Seaborn & Matplotlib

I am trying to create a Control Variability Grid Analysis in Python. This analysis requires plotting min and max blood glucose values (in this case, my own) per day on a scatterplot, with a background image that helps classify the overall glucose control. See here.

I have found other threads that address this question, but I cannot seem to get it to work for my application. I need the axes to be a specific range, and I think that is interfering with the image. The code I have tried so far, based on this page and this page, is as follows:

img = plt.imread('CVGA.png')
fig, ax = plt.subplots(figsize = (12,6))
sns.scatterplot(data=min_max_day,x=min_max_day['Glucose Value (mg/dL)']['amin'],
            y=min_max_day['Glucose Value (mg/dL)']['amax'], zorder=1)
plt.xlim(110,40)
plt.ylim(110,400)
ax.imshow(img);

And the output looks like this:

With the line “ax.imshow(img)” removed, the output looks like this:

Additional details:

• the image I’m using is 1200×1200:

• Maybe there is a problem with the fact that the scale I’m using isn’t a perfect square? Not sure if that makes sense, but clearly 110 to 400 on the Y axis and 100 to 40 on the X axis would not make a perfect square? I am unsure what to do about that.

Any advice or help would be greatly appreciated. Thanks a lot.

Answer

• Use the extent parameter as shown in Change values on matplotlib imshow() graph axis, but you must also use aspect='auto' and set figsize = (12, 12) (or (6, 6), etc.).
  • See all parameters at matplotlib.pyplot.imshow
• sns.scatterplot from the OP is commented out because no data was provided.
• Tested in python 3.8, matplotlib 3.4.2, and seaborn 0.11.1

import matplotlib.pyplot as plt
import seaborn as sns
import numpy as np  # sample data

img = plt.imread('CVGA.png')
fig, ax = plt.subplots(figsize=(6, 6))
# sns.scatterplot(data=min_max_day, x=min_max_day['Glucose Value (mg/dL)']['amin'], y=min_max_day['Glucose Value (mg/dL)']['amax'], zorder=1)

sns.scatterplot(x=np.linspace(110, 41, 10), y=np.linspace(110, 401, 10), ax=ax)
plt.xlim(110, 40)
plt.ylim(110, 400)

ax.imshow(img, extent=[110, 40, 110, 400], aspect='auto')