12

I'd like to create ellipses in matplotlib with a fill color that has an alpha (opacity) value that depends on the radius;

e.g., a 2D Gaussian.

Is there any way to do this?

It is possible to create rectangular plots with color gradients easily enough (like Gradient facecolor matplotlib bar plot and this) but I can't figure out how to do the same for circles/ellipses.

Community
  • 1
  • 1
keflavich
  • 18,278
  • 20
  • 86
  • 118

2 Answers2

8

Here is function example using the idea from Alex's post

import matplotlib.pyplot as plt,numpy as np

def gauplot(centers, radiuses, xr=None, yr=None):
        nx, ny = 1000.,1000.
        xgrid, ygrid = np.mgrid[xr[0]:xr[1]:(xr[1]-xr[0])/nx,yr[0]:yr[1]:(yr[1]-yr[0])/ny]
        im = xgrid*0 + np.nan
        xs = np.array([np.nan])
        ys = np.array([np.nan])
        fis = np.concatenate((np.linspace(-np.pi,np.pi,100), [np.nan]) )
        cmap = plt.cm.gray
        cmap.set_bad('white')
        thresh = 3
        for curcen,currad in zip(centers,radiuses):
                curim=(((xgrid-curcen[0])**2+(ygrid-curcen[1])**2)**.5)/currad*thresh
                im[curim<thresh]=np.exp(-.5*curim**2)[curim<thresh]
                xs = np.append(xs, curcen[0] + currad * np.cos(fis))
                ys = np.append(ys, curcen[1] + currad * np.sin(fis))
        plt.imshow(im.T, cmap=cmap, extent=xr+yr)
        plt.plot(xs, ys, 'r-')

And here is what you get when you run

    gauplot([(0,0), (2,3), (5,1), (6, 7), (6.1, 6.1)], [.3,. 4, .5, 1, .4], [-1,10], [-1,10])
             #           centers of circles           # radii of circles#

plot

sega_sai
  • 8,328
  • 1
  • 29
  • 38
  • I like the approach - I had considered it myself, but rejected it for a few reasons. 1. It doesn't work on log-log plots (not that patches work well, but they try) 2. It requires prior knowledge of the plot bounds (although this can be worked around) 3. it's pretty slow. Still, for the moment, I'll roll with it. – keflavich Jun 12 '12 at 17:04
  • 1
    I guess the origin of the imshow should be at the lower left corner, `plt.imshow(im.T, cmap=cmap, extent=xr+yr, origin="lower")`. – ImportanceOfBeingErnest Jul 14 '17 at 23:23
7

I don't think matplotlib currently supports gradient fills for patches - see this email.

john> Hello, I am trying to set a bar (a patched series of rectangles) with a fill pattern instead of just a solid color. Is there an easy way to do this in matplotlib?
john> I am thinking of something like Qt's QBrush which has cross, vertical, dense, etc. patterns.

There is no support for this currently -- it wouldn't be too hard to add for backends that support this kind of thing. Basically, we need to specify the API for it, and add support to backends. I have been wanting to add gradient fills for patches (eg polygons, rectangles) and it would be good to do both at once.

Instead of using patches you could create a mesh, calculate the colours with a function then use imshow with interpolation:

# Taken from http://matplotlib.sourceforge.net/examples/pylab_examples/layer_images.html

def func3(x,y):
    return (1- x/2 + x**5 + y**3)*exp(-x**2-y**2)

# make these smaller to increase the resolution
dx, dy = 0.05, 0.05

x = arange(-3.0, 3.0, dx)
y = arange(-3.0, 3.0, dy)
X,Y = meshgrid(x, y)

xmin, xmax, ymin, ymax = amin(x), amax(x), amin(y), amax(y)
extent = xmin, xmax, ymin, ymax

fig = plt.figure(frameon=False)

Z2 = func3(X, Y)

im2 = imshow(Z2, cmap=cm.jet, alpha=.9, interpolation='bilinear', extent=extent)

show()

This will result in the following (ignore the chequered background):

http://matplotlib.sourceforge.net/_images/layer_images.png

Alex L
  • 8,748
  • 5
  • 49
  • 75
  • and then putting a white mask z-intermediate with transparent patches where you want the markers, to get something more like gradient patches on a plain background. – cphlewis Apr 01 '15 at 17:33