I recommend using the griddata-method for interpolation. A sample would be:
import numpy as np
from matplotlib.mlab import griddata
import matplotlib.pyplot as plt
xs0 = np.random.random((1000)) * np.pi - np.pi/2
ys0 = np.random.random((1000)) * 3.5
zs0 = np.random.random((1000))
N = 30j
extent = (-np.pi/2,np.pi/2,0,3.5)
xs,ys = np.mgrid[extent[0]:extent[1]:N, extent[2]:extent[3]:N]
resampled = griddata(xs0, ys0, zs0, xs, ys)
plt.imshow(resampled.T, extent=extent)
plt.plot(xs0, ys0, "r.")
plt.plot(xs, ys, "b.")
plt.title("imshow for irregularly spaced data using griddata")
plt.show()
I guess transition from your 3*X-array to three X-arrays is obvious.
The result is:
Red points show the "original" positions of the data, blue points for the now regularly spaced data.
griddata returns a masked array. All points for which the interpolation cannot be evaluated are masked and then plotted as white areas.
HTH, Thorsten