Make diststats compatible with non-uniform P(r)

[luisfabib]

Most of the quantities calculated in diststats are based on moment analysis. However, the integration is mostly done using a simple summation

# Expectation operator function
def E(x,P):
    return np.sum(x*P/np.sum(P))

instead of trapezoidal integration, etc. This will return incorrect results when passing non-uniformly defined distance distributions.

Is not urgent, or difficult, but should be adapted to some point.