Is there some way I can color my scatter plot using two variables to set the color? ie. a sets the blue level and b sets the red level? Something along the lines of this:
import pylab
import numpy
x = numpy.random.random(50)
y = numpy.random.random(50)
a = numpy.random.random(50)
b = numpy.random.random(50)
s = pylab.scatter(x,y,c='red' * a + 'blue'*b)
pylab.colorbar(s)
pylab.show()
Expanding on falsetru's answer, you could make a two-dimensional colorbar using imshow:
import numpy as np
from matplotlib import pyplot as plt
x = np.random.random(50)
y = np.random.random(50)
a = np.random.random(50)
b = np.random.random(50)
s = c=zip(a, b, np.zeros(50))
ax = plt.gca()
print ax.get_position()
plt.scatter(x,y,c = s)
#adjust limits to make room for inset axes
plt.xlim(xmax = 1.5)
plt.ylim(ymax = 1.5)
#create inset axes
ax = plt.axes([.7, .675, .2, .2], axisbg='y')
n = 20
red = np.linspace(min(a), max(a), n)
green = np.linspace(min(b), max(b), n)
floats = np.linspace(0, 1, n)
#make arrays of all possible values between 0 and 1
X, Y = np.meshgrid(floats, floats)
#stack the arrays with a third array of zeros
Z = np.dstack((X, Y, np.zeros(X.shape)))
Z = np.rot90(Z)
red = np.round(red[::-1], 2)
green = np.round(green, 2)
plt.imshow(Z)
ticks = np.arange(0, n, 6)
plt.yticks(ticks, [red[i] for i in ticks])
plt.xticks(ticks, [green[i] for i in ticks], rotation = 'vertical')
plt.xlabel('Green')
plt.ylabel('Red')
plt.show()
pylab.scatter(x,y,c=zip(a, b, numpy.zeros(50)))?