1
$\begingroup$

This is a follow-up to the MathStackexchange question $3305$ and MathOverflow question $23229.$

The Bohr-Mollerup theorem states that the Gamma function is the unique function that satisfies:

$(1)\ f(x+1) = xf(x), \ (2)\ f(1) = 1, \ (3)\ \ln \circ f \ $ is convex.

Assume a function $f\colon \mathbb{R^{+}} \to \mathbb{R}$ that satisfies:

$(1)\ f(x+1) = xf(x), \ (2)\ f(1) = 1, \ (3)\ $ f is superadditive.

Is there a (natural) condition which makes this function unique?

Edit $1:$ The answer is 'no' as Moron explains. And if we assume $f$ superadditive only for $x,y \ge a$ for some real $a$?

Edit $2:$ I accept Moron's answer because it the correct answer to my question. Intended was the question in the sense of my first edit (and my first question). I am also curious to see an answer if condition $(3)$ reads: $(3'')$ $\ \ln \circ f \ $ is superadditive. Thanks to whuber for suggesting this.

  • 1
    This (a$n$d the $r$elated questions) would be far more interesting if you replaced "f is superadditive" by "ln(f) is superadditive".2010-08-26

1 Answers 1

4

There is no such function.

f(2) = f(1+1) = 1×f(1) = 1.

By superadditivity $f(1+1) \ge f(1) + f(1)$ i.e. $ 1 = f(2) \ge 2f(1) = 2$.

Not possible.