## The Radon-Nikodym Chain Rule

Today we take the Radon-Nikodym derivative and prove that it satisfies an analogue of the chain rule.

If , , and are totally -finite signed measures so that and , then -a.e. we have

By the linearity we showed last time, if this holds for the upper and lower variations of then it holds for itself, and so we may assume that is also a measure. We can further simplify by using Hahn decompositions with respect to both and , passing to subspaces on which each of our signed measures has a constant sign. We will from here on assume that and are (positive) measures, and the case where one (or the other, or both) has a constant negative sign has a similar proof.

Let’s also simplify things by writing

Since and are both non-negative there is also no loss of generality in assuming that and are everywhere non-negative.

So, let be an increasing sequence of non-negative simple functions converging pointwise to . Then monotone convergence tells us that

for every measurable . For every measurable set we find that

and so for all the simple we conclude that

Passing to the limit, we find that

and so the product serves as the Radon-Nikodym derivative of in terms of , and it’s uniquely defined -almost everywhere.

[…] Corollaries of the Chain Rule Today we’ll look at a couple corollaries of the Radon-Nikodym chain rule. […]

Pingback by Corollaries of the Chain Rule « The Unapologetic Mathematician | July 13, 2010 |

Could you quickly clarify why there is no loss of generality in assuming that f and g are everywhere non-negative? Sorry for the frequent recent posts, and thank you for the great resource too!

Comment by Bobby Brown | March 30, 2011 |

Basically, if they’re not you can always decompose into positive and negative parts, use the result there, and put everything back together.

Comment by John Armstrong | March 30, 2011 |

Could anyone tell me, were we exactly use the increasing sequences and why? I dont really see it :)

Comment by Wiebs91 | November 16, 2012 |

You mean where we use the fact that the sequence is increasing? That’s a requirement of the monotone convergence theorem.

Comment by John Armstrong | November 16, 2012 |

I know that theorem. I just wanted to know why we need it here, where it is used in the last part V(E)=… . I’m sorry, im really not used to this notations at all and don’t get the point of it by now

Comment by Wiebs91 | November 16, 2012 |

Okay, well the point is that any measurable function can be approximated as the limit of an increasing sequence of measurable functions , and the simple functions are basically constants times characteristic functions. So what we do is prove our result for the characteristic functions . Then the fact that everything in sight is linear means that it holds for simple functions. And finally we can pass to the limit (using monotone convergence) and get the result for all measurable functions.

Comment by John Armstrong | November 16, 2012 |

Ah, now I see it, thanks for your time an help :)

Comment by Wiebs91 | November 16, 2012 |