Matplotlib: converting a matrix to a raster image¶
| Date: | 2009-02-26 (last modified), 2006-07-10 (created) |
|---|
Scipy provides a command (imsave) to make a raster (png, jpg...) image from a 2D array, each pixel corresponding to one value of the array. Yet the image is black and white.
Here is a recipy to do this with Matplotlib, and use a colormap to give color to the image.
In [1]:
from matplotlib.pyplot import *
from scipy import mgrid
def imsave(filename, X, **kwargs):
""" Homebrewed imsave to have nice colors... """
figsize=(array(X.shape)/100.0)[::-1]
rcParams.update({'figure.figsize':figsize})
fig = figure(figsize=figsize)
axes([0,0,1,1]) # Make the plot occupy the whole canvas
axis('off')
fig.set_size_inches(figsize)
imshow(X,origin='lower', **kwargs)
savefig(filename, facecolor='black', edgecolor='black', dpi=100)
close(fig)
X,Y=mgrid[-5:5:0.1,-5:5:0.1]
Z=sin(X**2+Y**2+1e-4)/(X**2+Y**2+1e-4) # Create the data to be plotted
imsave('imsave.png', Z, cmap=cm.hot )
In [2]:
imshow(imread('imsave.png'))
Out[2]:
Section author: GaelVaroquaux, Unknown[123], Unknown[124]
Attachments
