Showing $k! \cdot q_k(n) \leq \binom{n-1}{k-1}$?

65 Views Asked by At

Let $q_k(n)$ denote the number of distinct partitions of some $n \in \mathbb{N}$ with exactly $k$ parts. How can I show that: $$k! \cdot q_k(n) \leq \binom{n-1}{k-1}$$

There is a related inequality that I reasoned out which was $k! \cdot p_k(n) \geq \binom{n-1}{k-1}$, where $p_k(n)$ denotes the number of partitions of $n$ with exactly $k$ parts. The way I justified this one was by recognizing the right-hand side counts the number of ordered $k$-tuples of positive integer solutions $(x_1, x_2, \ldots, x_k)$ satisfying $$\sum_{i=1}^k x_i = n$$

By definition, $p_k(n)$ is the number of partitions of $n$ with $k$ parts, and order is not considered within a partition. By multiplying this by $k!$ we count each permutation of the $k$-sized partitions.

For some ordered $k$-tuple of positive integer solutions, there are at most $k!$ distinct $k$-tuples containing the same numbers. This only occurs when all $k$ numbers are distinct. The inequality $k! \cdot p_k(n) \geq \binom{n-1}{k-1}$ follows.

Is there some similar reasoning to reach the first inequality?

1

There are 1 best solutions below

1
On BEST ANSWER

You can use essentially the same argument as in your proof for the related inequality. Each partition with distinct parts corresponds to exactly $k!$ distinct compositions of the number $n$ with $k$ parts. So there are at least $k! q_k(n)$ compositions of $n$ with $k$ parts, and as in your proof for the related inequality, we know that the number of compositions of $n$ with $k$ parts is $$ \binom{n - 1}{k - 1}. $$

i.e. If $x_1 < x_2 < \dots < x_k$ is such that $x_1 + x_2 + \dots + x_k = n$, so that $(x_1, x_2, \dots, x_k)$ is one of the partitions counted by $q_k(n)$, then this corresponds to $k!$ solutions to $$ y_1 + y_2 + \dots + y_k = n $$ because we can take $y_i = x_{\sigma(i)}$ where $\sigma$ is some permutation of $\{1, 2, \dots, k\}$, and there are $k!$ such permutations. This implies that there are at least $k! q_k(n)$ solutions to $$ y_1 + y_2 + \dots + y_k = n $$ but we know that there are exactly $$ \binom{n - 1}{k - 1} $$ solutions.