Resizing imshow heatmap into a given image size in matplotlib

Question

I'm using matplotlib for plotting, and I have the following matrix, Mat which I would like to plot into a heatmap.

Mat.shape

which yields (20,20).

I use the following code to plot it into a heatmap, following the this.

plt.imshow(Mat, cmap='Reds', interpolation='nearest')    
plt.show()

But I have to resize this heatmap into 1600 x 1200(x,y) size, since, I want it to be overlapped with the image, img. The code is as follows.

plt.imshow(img, alpha=.5) # for image
plt.xlim(0, 1600)
plt.ylim(1200, 0)
plt.axis('off')

plt.imshow(Mat, cmap='Reds', interpolation='nearest', alpha=.5)   # for heatmap to overlap
plt.show()

For some reason I would not like to change the size of img.

---

My try

I tried to resize at the level of plt.imshow(Mat), which I found hard. The only option I see at present is to resize Mat into 1600 x 1200 matrix with redundancy. Anyone to give me some efficient solution?

Answer 1

Use the extent= parameter of imshow to scale your matrix to the image's scale (more information here as well).

plt.figure()
img = plt.imread('stinkbug.png')
plt.imshow(img, alpha=1) # for image
#plt.axis('off')

xmin, xmax = plt.xlim()
ymin, ymax = plt.ylim()

Mat = np.random.normal(size=(20,20))
plt.imshow(Mat, cmap='Reds', interpolation='nearest', alpha=.5, extent=(xmin,xmax,ymin,ymax))   # for heatmap to overlap
plt.show()