Skip to content

Commit

Permalink
Small grammar fixes
Browse files Browse the repository at this point in the history
  • Loading branch information
ccodel committed Aug 22, 2024
1 parent 197b866 commit cc6af5f
Show file tree
Hide file tree
Showing 2 changed files with 4 additions and 4 deletions.
2 changes: 1 addition & 1 deletion ITP/symmetry-breaking.tex
Original file line number Diff line number Diff line change
Expand Up @@ -11,7 +11,7 @@
there exists a list of points in \emph{canonical position} with the same triple-orientations.
Canonical position is defined as follows.
\begin{definition}[Canonical Position]
A list of points~$L = (p_1,\ldots, p_{n})$ is said to be in \emph{canonical position} if it satisfies all the following properties:
A list of points~$L = (p_1,\ldots, p_{n})$ is said to be in \emph{canonical position} if it satisfies all of the following properties:
\begin{itemize}
\item \textbf{(General Position)} No three points are collinear, i.e., for all $1 \leq i < j < k \leq n$, we have $\sigma(p_i, p_j, p_k) \neq 0$.
\item \textbf{($x$-order)} The points are sorted with respect to their $x$-coordinates, i.e., $x(p_i) < x(p_j)$ for all $1 \leq i < j \leq n$.
Expand Down
6 changes: 3 additions & 3 deletions ITP/triple-orientations.tex
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
problems concerning the existence of an object~$\mathcal{O}$ in a continuous search space like $\mathbb{R}^2$
must be reformulated in terms of the existence of a discrete, finitely-representable object~$\mathcal{O}'$
that a computer can search for.
It is especially challenging to discretize problems in which the desired geometric object $\mathcal{O}$ is characterized by very specific coordinates of points,
It is especially challenging to discretize problems for which the desired geometric object~$\mathcal{O}$ is characterized by very specific coordinates of points,
thus requiring the computer to use floating-point arithmetic,
which suffers from numerical instability.
Fortunately,
Expand Down Expand Up @@ -177,7 +177,7 @@
% Let us now discuss the previous steps. First,~\lstinline|(PtInTriangle a p q r)| presents a \emph{ground-truth} definition for membership in a triangle, in terms of~\textsf{mathlib}'s \lstinline|ConvexHull| definition, whereas~\lstinline|(σPtInTriangle a p q r)| is based on orientations.
Heule and Scheucher used the orientation-based definition~\cite{emptyHexagonNumber}, and as it is common in the area, its equivalence to the \emph{ground-truth} mathematical definition was left implicit.
Heule and Scheucher used the orientation-based definition~\cite{emptyHexagonNumber}, and, as is common in the area, its equivalence to the \emph{ground-truth} mathematical definition was left implicit.
This equivalence, formalized in~\lstinline|theorem σPtInTriangle_iff| is not trivial to prove:
the forward direction in particular requires reasoning about convex combinations and determinants.
Using the previous theorem, we can generalize to $k$-gons as follows.
Expand Down Expand Up @@ -214,7 +214,7 @@
SAT encodings for Erd\H{o}s-Szekeres-type problems use triple orientations to capture properties like convexity and emptiness,
thus discretizing the problem.
Because we have proved that $\pi_k(S)$ is an orientation property,
the values of $\sigma$ on the points in $S$ contain enough information to determine whether $\pi_k(S)$.
the values of $\sigma$ on the points in $S$ contain enough information to determine whether $\pi_k(S)$ holds.
Therefore, we have proved that given $n$ points,
it is enough to analyze the values of $\sigma$ over these points,
a discrete space with at most $2^{n^3}$ possibilities,
Expand Down

0 comments on commit cc6af5f

Please sign in to comment.