PDA

View Full Version : Suggestion for proof


Yoss
Sep23-04, 02:20 PM
Let a and b be natural numbers and LCM(a,b) = m.
Prove that if GCD(a,b) = 1, then LCM(a,b) = ab.

What I got so far was that since b|LCM(a,b), then b*k = LCM(a,b) for some natural number k. I know I need to show that k </= a and k >/= a so I get a = k. And show that b*k = ba = LCM(a,b).

I can get a </= k:

Since LCM(a,b) = m, then a|m and b|m. Since LCM(a,b) = m = b*k, then a|bk. <And by the Euclid's Lemma, if a|bc and GCD(a,b) = 1, then a|c>. So
a|k. So a*j = k for some natural number j, and therefore a </= k.

I'm sure I have to derive that a >/= k by the antecedent. I'm trying to use the property that GCD(a,b) = 1 = a*x + b*y for some integers x and y by showing that k|a. But I'm stuck.

Any suggestions? Thanks

Edit: Maybe I should have posted this in Number Theory. Mod, can you move it if you think it should be there?

NateTG
Sep23-04, 03:09 PM
There's a nice proof using the fundemental theorem of algebra.

Yoss
Sep23-04, 03:18 PM
There's a nice proof using the fundemental theorem of algebra.

I'm sorry NateTG, I don't follow. Can you please elaborate?

Yoss
Sep23-04, 04:20 PM
I think you might mean Fundamental Theorem of Arithmetic.

NateTG
Sep23-04, 04:57 PM
I think you might mean Fundamental Theorem of Arithmetic.
Yep. Sorry. Apparently I need to sleep more.

Yoss
Sep23-04, 05:14 PM
Am I even on the right track? It feels like I'm missing some trivial property that I need to solve this. Can anyone suggest something, and perhaps how to apply the FTA to this? Thanks.

NateTG
Sep23-04, 07:09 PM
Yeah, I think you're almost there.
You're trying to prove that a \geq k, right?
You might want to use the fact that LCM(a,b) \leq ab since ab is obviously a common multiple.

Gokul43201
Sep23-04, 08:25 PM
Also, the more general result :(a,b)*((a,b)) = ab where () is GCD, (( )) is LCM follows directly from

(a,b) = \prod_i p_i^{min(k_i,l_i)} ~~ and

((a,b)) = \prod_i p_i^{max(k_i,l_i)}~~ where

a = \prod_i p_i^{k_i} ~~ and

b = \prod_i p_i^{l_i}

Yoss
Sep24-04, 12:19 AM
Yeah, I think you're almost there.
You're trying to prove that a \geq k, right?
You might want to use the fact that LCM(a,b) \leq ab since ab is obviously a common multiple.

I knew it was something trivial I didn't see. Thanks NateG and Goku.