from scipy.interpolate import interp1d

x = np.linspace(0, 10, 10)
y = np.exp(-x/3.0)
f = interp1d(x, y)
f2 = interp1d(x, y, kind='cubic')

xnew = np.linspace(0, 10, 40)
import matplotlib.pyplot as plt
plt.plot(x,y,'o',xnew,f(xnew),'-', xnew, f2(xnew),'--')
plt.legend(['data', 'linear', 'cubic'], loc='best')
plt.show()