Plot with simple legend

Use 'label' in your plot() call.

import math
import matplotlib.pyplot as plt

xv= map( lambda x: (x/4.)-10., range(0,81))

for l in [ 0.1, 0.5, 1., 5.] :
    yv= map( lambda x: math.exp((-(-x)**2)/l), xv)
    plt.plot(xv,yv,label='lambda = '+str(l));

plt.legend() 
plt.show()

Sidenote: the function plotted is that of the Gaussian kernel in weighted nearest neighour regression, with xi=0