Bounded section of pointed cone
Dependencies:
- Polyhedral set and polyhedral cone
- Rank of a matrix
- Condition for polyhedral cone to be pointed
- Rank of a homogenous system of linear equations
- Direction of a convex set and Recession cone
- Polyhedron is unbounded iff it has direction
- Recession cone of a polyhedron
$\newcommand{\defeq}{=}$ $\newcommand{\xhat}{\widehat{x}}$ $\newcommand{\rank}{\operatorname{rank}}$ Let $C \defeq \{x \in \mathbb{R}^n: (a_i^Tx \ge 0, \forall i \in I) \textrm{ and } (a_i^Tx = 0, \forall i \in E)\}$ be a pointed cone. Let $A$ be the matrix whose rows are $\{a_i: i \in I \cup E\}$. Then $\rank(A) = n$.
Let $K \subseteq I \cup E$ and let $B$ be the matrix having rows $\{a_i: i \in K\}$ such that $\rank(B) = n$. Let $b \defeq \sum_{i \in I \cup E} \lambda_ia_i$, where $\lambda_i > 0$ for $i \in K$. Let $\gamma > 0$ and let $P = \{x \in C: b^Tx = \gamma\}$. Then $P$ is bounded and $b^Tx > 0$ for any $x \in C - \{0\}$. Hence, $\forall x \in C - \{0\}$, $(\gamma/b^Tx)x \in P$.
Proof
Let $\xhat \in C - \{0\}$.
Lemma 1: $\exists i \in K$ such that $a_i^T\xhat > 0$.
Proof: Suppose $a_i^T\xhat = 0$ for all $i \in K$. Then $B\xhat = 0$. Hence, $\xhat = 0$ since $\rank(B) = n$, which is a contradiction. Hence, $\exists i \in K$ such that $a_i^T\xhat > 0$. □
Lemma 2: $b^T\xhat > 0$ and $(\gamma/b^T\xhat)\xhat \in P$.
Proof: Suppose $a_k^T\xhat > 0$, where $k \in K$ (by Lemma 1). Then $b^T\xhat = \sum_{i \in I \cup E} \lambda_i a_i^T\xhat \ge \lambda_k a_k^T\xhat > 0$. Let $\gamma/b^T\xhat)$. Then $\mu > 0$, so $\mu\xhat \in C$. $b^T(\mu\xhat) = \gamma$, so $\mu\xhat \in P$. □
Lemma 3: $P$ is bounded.
Proof: Suppose $P$ is not bounded. Then $P$ has a non-zero direction $d$. Hence, $d \in C-\{0\}$ and $b^Td = 0$. Since $d \in C$, we get $Bd \ge 0$. So, $a_i^Td \ge 0$ for $i \in K$. Since $0 = b^Td = \sum_{i \in I \cup E} \lambda_i(a_i^Td) \ge \sum_{i \in K} \lambda_i(a_i^Td) \ge 0$ and $\lambda_i > 0$ for $i \in K$, we get that $a_i^Td = 0$ for all $i \in K$. Hence, $Bd = 0$. But since $\rank(B) = n$, this means $d = 0$, which is a contradiction. □
Dependency for:
Info:
- Depth: 12
- Number of transitive dependencies: 52
Transitive dependencies:
- /linear-algebra/vector-spaces/condition-for-subspace
- /linear-algebra/matrices/gauss-jordan-algo
- /sets-and-relations/equivalence-relation
- Group
- Ring
- Polynomial
- Integral Domain
- Comparing coefficients of a polynomial with disjoint variables
- Field
- Vector Space
- Linear independence
- Span
- Semiring
- Matrix
- Stacking
- System of linear equations
- Product of stacked matrices
- Matrix multiplication is associative
- Reduced Row Echelon Form (RREF)
- Matrices over a field form a vector space
- Row space
- Elementary row operation
- Every elementary row operation has a unique inverse
- Row equivalence of matrices
- Row equivalent matrices have the same row space
- RREF is unique
- Identity matrix
- Inverse of a matrix
- Inverse of product
- Elementary row operation is matrix pre-multiplication
- Row equivalence matrix
- Equations with row equivalent matrices have the same solution set
- Rank of a matrix
- Basis of a vector space
- Linearly independent set is not bigger than a span
- Homogeneous linear equations with more variables than equations
- Rank of a homogenous system of linear equations
- Cone
- Convex combination and convex hull
- Transitivity of convexity
- Convex set
- Polyhedral set and polyhedral cone
- Basic feasible solutions
- Condition for existence of BFS in a polyhedron
- Point in polytope is convex combination of BFS
- Convex hull of a finite number of points in Euclidean space is bounded
- Extreme point of a convex set
- Condition for a point to be extreme
- Condition for polyhedral cone to be pointed
- Direction of a convex set and Recession cone
- Recession cone of a polyhedron
- Polyhedron is unbounded iff it has direction