[Bf-committers] Matrix "differencing" procedure

[Martin Poirier]

--- Chris Want <cwant@ualberta.ca> wrote:

> Martin Poirier wrote:
> > RA = B
> > 
> > You're looking for R.
> > 
> > RiAA = BiA (where iA is inverse of A)
> 
> Blindly stuffing iA in the middle like that leads to
> a really bad illustration, since it seems to imply
> that
> matrix multiplication is commutitive (i.e.,
> B*C = C*B, which is not generally true).
> If this were a linear algebra class you would
> definitely lose marks.
> 
> The better illustration would be RAiA = BiA,
> or more clearly, as (RA)iA = (B)iA.
> i.e., what happened was that both sides were
> multiplied on the right by iA.
> 
> Chris <-- nitpicker, but it's important to get it
> straight

Good eye. 
That's what I get for not double checking after
changing stuff (at first, I was doing AR = B -> iAAR =
iAB and decided to switch that around).

That's not nitpicking, it's called being precise.

Lets rewrite that as a whole with correct
associativity parenthesis:

RA = B

(RA)iA = BiA

R(AiA) = BiA

R(I) = BiA

R = BiA

Martin


		
__________________________________
Do you Yahoo!?
Yahoo! Mail - 50x more storage than other providers!
http://promotions.yahoo.com/new_mail