Prove or disprove: $\mathcal{I}$ is the collection of independent sets of a matroid.

372 Views Asked by At

Let $\nu_1,...,\nu_n \in \mathbb{R}^n$ be vectors. Let $\mathcal{I}$ be the collection of subsets $I \subseteq \{1,2,...,k\}$ for which the $\nu_i, i\in I$ are affinely independent.

Exercise: I need to prove or disprove that $\mathcal{I}$ is the collection of independent sets of a matroid.

I know that the pair $M = (E,\mathcal{I})$ is a matroid if the properties

i) $\emptyset \in \mathcal{I}$.

ii) If $I\subseteq J$ and $J\in\mathcal{I}$, then $I\in\mathcal{I}$.

iii) If $I,J\in \mathcal{I}$ and $\left|I\right| <\left|J\right|$, then there is an element $e\in J\backslash I$ such that $I+e\in\mathcal{I}$.

all hold. In this case $I \in \mathcal{I}$ are called the independent sets.

So in order to solve the exercise I need to show that properties i),ii) and iii) all hold if $\mathcal{I}$ is the collection of subsets $I \subseteq \{1,2,...,k\}$ for which the $\nu_i, i\in I$ are affinely independent.

What I've tried: I had to look up affine independence for this exercise and the most comprehensible definition I came across was this one:

The points $x_k$ are affine independent when $$ \sum \lambda_k x_k = 0 \text{ with }\sum \lambda_k =0 $$ implies all $\lambda_k = 0$.

I have no idea how to use this definition though. Let's say I want to show that $\emptyset\in \mathcal{I}$ (or $\emptyset \notin\mathcal{I}$). How do I show that $\emptyset$ is affinely independent or affinely dependent? And if I need to show that several vectors are affinely independent, do I look at all the $\sum\lambda_k\nu_k$? Since points in $\mathbb{R}^n$ are given by vectors right?

Questions:

  • How do I use affine independence to show that $\emptyset\in\mathcal{I}$ or $\emptyset\notin\mathcal{I}$?

  • Take vectors $\nu_i$ for some $I$. How do I show that $\nu_i$ are affinely (in)dependent?

  • (optional) How do I solve this exercise?

Thanks in advance!

1

There are 1 best solutions below

0
On BEST ANSWER
  • How do I use affine independence to show that $\emptyset\in\mathcal{I}$ or $\emptyset\notin\mathcal{I}$?

When you're working with the empty set, the two sums in your definition of affine independence are the empty sum. By convention, this is zero.

  • Take vectors $\nu_i$ for some $I$. How do I show that $\nu_i$ are affinely (in)dependent?

You don't strictly need this to do the exercise, but it's a good question if you want to play with some examples. If $M$ is the matrix containing your affine vectors as columns, then $M$ is affinely independent if and only if $M$ with a row of all $1$s is linearly independent. (It's a small exercise to show that this is equivalent to the definition that you gave.) So you can use the same techniques that you use to show linear (in)dependence.