You are on page 1of 3

PMath 451/651, Fall Term 2012 Homework Assignment 6

Posted on Saturday, November 3; due on Friday, November 9

Problem 1. Let (X, M, ) be a measure space, and let f, g be two functions in L1 (). (a) Suppose that f = g a.e. (in the sense dened in Problem 2 of Assignment 5). Prove that A f d = A g d for all A M. (b) Conversely, suppose that Solution.

d =

A g d

for all A M. Prove that f = g a.e..

Problem 2 is about Lebesgue-Stieltjes measures on the Borel -algebra BR of the real line. If : BR [0, ] is a Lebesgue-Stieltjes measure and if G is the centered Stieltjes function associated to , it is customary (especially in probability textbooks) to use the notation f (t)d G(t) instead of f (t)d (t), for a function f L1 (). This notation can be further shortened by saying (somewhat abusively) that we have d (t) = d G(t). We will denote by the Lebesgue measure on R. The centered Stieltjes function of is the identity function, hence the shortened notation from the preceding paragraph comes in this case to d (t) = dt. With all these shortened notations, Problem 2 can be summarized by the formula d G(t) = G (t) dt.

Problem 2. Let G : R R be a dierentiable non-decreasing function such that G(0) = 0. We denote G =: g. On the other hand let : BR [0, ] be the LebesgueStieltjes measure which is uniquely determined by the requirement that the centered Stieltjes function of is equal to G. (a) Prove that g is a non-negative Borel function. (b) Let : BR [0, ] be the positive measure dened (in the sense of Lecture 11) by the formula d (t) = g(t)d (t), where is the Lebesgue measure on R. That is, we put (A) := A g d, A BR . Prove that = . Solution.

In Problem 3 we x a measurable space (X, M) and we consider the space of nite signed measures Meas (X, M) that was introduced in Lecture 11. Recall that Meas (X, M) is a vector space over R, with operations dened in the natural way: (1 1 + 2 2 )(A) := 1 1 (A) + 2 2 (A), for every 1 , 2 Meas (X, M) , 1 , 2 R , A M. It will be convenient to also use the partial order on Meas (X, M) dened as follows: for 1 , 2 Meas (X, M) we write 1 2 to mean that 1 (A) 2 (A) for every A M. Equivalently, one can say that 1 2 2 1 Meas+ (X, M) , where Meas+ (X, M) is the subset of Meas (X, M) consisting of positive nite measures.

Problem 3. (a) Let be in Meas (X, M), and consider the total variation measure || Meas+ (X, M). Prove that for every A M one has ||(A) = sup |(B)| + |(C)| | B, C M, B, C A, B C = . (b) Let be a measure in Meas (X, M). Suppose is a measure in Meas+ (X, M) with the property that (A) |(A)|, A M. Prove that ||. (c) Let be a measure in Meas (X, M) and let be a real number. Prove that | | = || ||. (d) Let 1 , 2 be two measures in Meas (X, M). Prove that |1 + 2 | |1 | + |2 |. (e) For every Meas (X, M), we denote |||| := ||(X). Prove that |||| is a norm on the vector space Meas (X, M). Solution.