MathOverflow is a question and answer site for professional mathematicians. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Hello all, the $\Delta$ operator on functions $\mathcal{N} \to \mathbb{R}$ (where $\mathcal N$ denotes $\lbrace 1,2, \ldots , \rbrace$ )defined by $\Delta(f)(n)=f(n+1)-f(n)$ is well-known and it is not very hard to show by induction that $f$ is a polynomial of degree $\leq k$ iff $\Delta^{k+1}(f)$ is identically zero, where $\Delta^{k+1}$ denotes $\Delta$ iterated $k+1$ times. Now I say that a function $f : \mathcal{N} \to \mathbb{R}$ is "almost polynomial" iff $\Delta^{k}(f)$ is a bounded function for some $k$.

My question is : let $\lambda >0$ be a non-integer, and let $f(n)=n^{\lambda}$. Is it true that $f$ is almost polynomial ?

share|cite|improve this question
Should that be $f(n)=n^\lambda$ in the question? – Bruce Westbury Mar 13 '10 at 11:23
Strange. The $\lambda$ as exponent has now appeared. – Bruce Westbury Mar 13 '10 at 11:25
up vote 9 down vote accepted

One can also get this from a standard form of the remainder in Taylor's theorem. Namely, if $k>\lambda$, and $T_{k-1}(x)$ is the degree $k-1$ Taylor polynomial of $f(x)$ at $x=n$, then $$f(x) = T_{k-1}(x) + \frac{f^{(k)}(\xi_x)}{k!} (x-n)^k$$ for some $\xi_x \in [n,x]$. Applying $\Delta^k$ kills the Taylor polynomial term, and when we evaluate the result at $x=n$, the linear combination of the remainders is small when $n$ is large relative to $k$, because each $f^{(k)}(\xi_x)$ is $O(n^{\lambda-k})$ as $n \to \infty$.

share|cite|improve this answer
Thanks Bjorn. Your solution is simpler. – Ewan Delanoy Mar 14 '10 at 5:27

There is an "iterated" version of the mean value theorem which states that for a smooth enough function $f$ on an interval that $\Delta^k f(x)=f^{(k)}(\xi)$ where $\xi$ is between $x$ and $x+k$. This should be enough for your purposes.

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.