Computational Physics - Department of Physics

11.2 Probability Distribution Functions 351

11.1.6Brief Summary.

In essence the Monte Carlo method contains the following ingredients

• A PDF which characterizes the system


• Random numbers which are generated so as to cover in an as uniform as possible
  way on the unity interval [0,1].


• A sampling rule


• An error estimation


• Techniques for improving the errors

In the next section we discuss various PDF’s which may be of relevance here, thereafter
we discuss how to compute random numbers. Section 11.4 discusses Monte Carlo integration
in general, how to choose the correct weighting function andhow to evaluate integrals with
dimensionsd> 1.

11.2 Probability Distribution Functions.

Hitherto, we have tacitly used properties of probability distribution functions in our computa-
tion of expectation values. Here and there we have referred to the uniform PDF. It is now time
to present some general features of PDFs which we may encounter when doing physics and
how we define various expectation values. In addition, we derive the central limit theorem
and discuss its meaning in the light of properties of variousPDFs.
The following table collects properties of probability distribution functions. In our notation
we reserve the labelp(x)for the probability of a certain event, whileP(x)is the cumulative
probability.

Table 11.2Important properties of PDFs.
Discrete PDF Continuous PDF
Domain {x 1 ,x 2 ,x 3 ,...,xN} [a,b]
Probability p(xi) p(x)dx
Cumulative Pi=∑il= 1 p(xl) P(x) =∫axp(t)dt
Positivity 0 ≤p(xi)≤ 1 p(x)≥ 0
Positivity 0 ≤Pi≤ 1 0 ≤P(x)≤ 1
Monotonic Pi≥Pjifxi≥xj P(xi)≥P(xj)ifxi≥xj
Normalization PN= 1 P(b) = 1

With a PDF we can compute expectation values of selected quantities such as

〈xk〉=^1

N

N

∑

i= 1

xkip(xi),

if we have a discrete PDF or
〈xk〉=

∫b

a

xkp(x)dx,

in the case of a continuous PDF. We have already defined the mean valueμand the variance
σ^2.