how to use 'extent' in matplotlib.pyplot.imshow

Here's an example based on http://matplotlib.org/examples/pylab_examples/image_demo3.html showing use of extent.

#!/usr/bin/env python
from pylab import *
try:
    from PIL import Image
except ImportError, exc:
    raise SystemExit("PIL must be installed to run this example")

import matplotlib.cbook as cbook

datafile = cbook.get_sample_data('ada.png')
h = Image.open(datafile)
dpi = rcParams['figure.dpi']
figsize = h.size[0]/dpi, h.size[1]/dpi

figure(figsize=figsize)
ax = axes([0,0,1,1], frameon=False)
ax.set_axis_off()
ax.set_xlim(0,2)
ax.set_ylim(0,2)
im = imshow(h, origin='upper',extent=[-2,4,-2,4])  # axes zoom in on portion of image
im2 = imshow(h, origin='upper',extent=[0,.5,0,.5]) # image is a small inset on axes

show()

If you don't set your axis limits, they become your extents & then don't seem to have any effect.

Specify, in the coordinates of your current axis, the corners of the rectangle that you want the image to be pasted over

Extent defines the left and right limits, and the bottom and top limits. It takes four values like so: extent=[horizontal_min,horizontal_max,vertical_min,vertical_max].

Assuming you have longitude along the horizontal axis, then use extent=[longitude_top_left,longitude_top_right,latitude_bottom_left,latitude_top_left]. longitude_top_left and longitude_bottom_left should be the same, latitude_top_left and latitude_top_right should be the same, and the values within these pairs are interchangeable.

If your first element of your image should be plotted in the lower left, then use the origin='lower' imshow option as well, otherwise the 'upper' default is what you want.