Prove that if p is a prime number, then p divides (np)−⌊np⌋, for all n>p.
(where the ⌊np⌋ denotes the greatest integer less than or equal to np, for any real number np.)
Does this result generalise to a result about pr instead of p?, where r=np
My only thought on approaching this is assuming that it is true and proving my contradiction, but am unsure on how to use the floor function as it can give the same result for many numbers, for example 4.1,4.2,4.4,4.4 all output 4.
Any help would be appreciated.
Answer
Lucas' theorem does the job. Write n=q∑0pknk where nk's are the digits in the base p-representation of n. Since n>p, we must have q≥1. Then, by Lucas' theorem,
(np)≡(n11)(n00)≡n1(modp)
Also, ⌊n/p⌋=⌊(q∑0pknk)/p⌋=⌊n0/p⌋+n1+p(⋯)=0+n1+p(⋯)≡n1(modp)
Thus, (np)≡⌊np⌋(modp)
I'm not sure if an analogous argument would work for modulo prime powers of p because Lucas' theorem modulo prime powers is a bit different from the usual statement, but here's a (sort of) generalization:
(npk)≡⌊npk⌋≡nk(modp)
where k goes from 1 to q (also the trivial case of k=0 holds)
No comments:
Post a Comment