Let $X$ be a set.
-
Lack of Functoriality. The assignment $\webleft (U,V\webright )\mapsto U\mathbin {\triangle }V$ does not in general define functors
\[ \begin{array}{ccc} U\mathbin {\triangle }-\colon \mkern -15mu & \webleft (\mathcal{P}\webleft (X\webright ),\subset \webright ) \mkern -17.5mu& {}\mathbin {\to }\webleft (\mathcal{P}\webleft (X\webright ),\subset \webright ),\\ -\mathbin {\triangle }V\colon \mkern -15mu & \webleft (\mathcal{P}\webleft (X\webright ),\subset \webright ) \mkern -17.5mu& {}\mathbin {\to }\webleft (\mathcal{P}\webleft (X\webright ),\subset \webright ),\\ -_{1}\mathbin {\triangle }-_{2}\colon \mkern -15mu & \webleft (\mathcal{P}\webleft (X\webright )\times \mathcal{P}\webleft (X\webright ),\subset \times \subset \webright ) \mkern -17.5mu& {}\mathbin {\to }\webleft (\mathcal{P}\webleft (X\webright ),\subset \webright ). \end{array} \]
-
Via Unions and Intersections. We have
\[ U\mathbin {\triangle }V=\webleft (U\cup V\webright )\setminus \webleft (U\cap V\webright ) \]
for each $U,V\in \mathcal{P}\webleft (X\webright )$, as in the Venn diagram
-
Symmetric Differences of Disjoint Sets. If $U$ and $V$ are disjoint, then we have
\[ U\mathbin {\triangle }V=U\cup V. \]
-
Associativity. The diagram
commutes, i.e. we have
\[ \webleft (U\mathbin {\triangle }V\webright )\mathbin {\triangle }W = U\mathbin {\triangle }\webleft (V\mathbin {\triangle }W\webright ) \]
for each $U,V,W\in \mathcal{P}\webleft (X\webright )$, as in the Venn diagram
-
Unitality. The diagrams commute, i.e. we have
\begin{align*} U\mathbin {\triangle }\text{Ø}& = U,\\ \text{Ø}\mathbin {\triangle }U & = U \end{align*}
for each $U\in \mathcal{P}\webleft (X\webright )$.
-
Commutativity. The diagram
commutes, i.e. we have
\[ U\mathbin {\triangle }V = V\mathbin {\triangle }U \]
for each $U,V\in \mathcal{P}\webleft (X\webright )$.
-
Invertibility. We have
\[ U\mathbin {\triangle }U = \text{Ø} \]
for each $U\in \mathcal{P}\webleft (X\webright )$.
-
Interaction With Unions. We have
\[ \webleft (U\mathbin {\triangle }V\webright )\cup \webleft (V\mathbin {\triangle }T\webright ) = \webleft (U\cup V\cup W\webright )\setminus \webleft (U\cap V\cap W\webright ) \]
for each $U,V,W\in \mathcal{P}\webleft (X\webright )$.
-
Interaction With Complements I. We have
\[ U\mathbin {\triangle }U^{\textsf{c}}= X \]
for each $U\in \mathcal{P}\webleft (X\webright )$.
-
Interaction With Complements II. We have
\begin{align*} U\mathbin {\triangle }X & = U^{\textsf{c}},\\ X\mathbin {\triangle }U & = U^{\textsf{c}}\end{align*}
for each $U\in \mathcal{P}\webleft (X\webright )$.
-
Interaction With Complements III. The diagram
commutes, i.e. we have
\[ U^{\textsf{c}}\mathbin {\triangle }V^{\textsf{c}}=U\mathbin {\triangle }V \]
for each $U,V\in \mathcal{P}\webleft (X\webright )$.
-
“Transitivity”. We have
\[ \webleft (U\mathbin {\triangle }V\webright )\mathbin {\triangle }\webleft (V\mathbin {\triangle }W\webright )=U\mathbin {\triangle }W \]
for each $U,V,W\in \mathcal{P}\webleft (X\webright )$.
-
The Triangle Inequality for Symmetric Differences. We have
\[ U\mathbin {\triangle }W\subset U\mathbin {\triangle }V\cup V\mathbin {\triangle }W \]
for each $U,V,W\in \mathcal{P}\webleft (X\webright )$.
-
Distributivity Over Intersections. We have
\begin{align*} U\cap \webleft (V\mathbin {\triangle }W\webright ) & = \webleft (U\cap V\webright )\mathbin {\triangle }\webleft (U\cap W\webright ),\\ \webleft (U\mathbin {\triangle }V\webright )\cap W & = \webleft (U\cap W\webright )\mathbin {\triangle }\webleft (V\cap W\webright ) \end{align*}
for each $U,V,W\in \mathcal{P}\webleft (X\webright )$.
-
Interaction With Characteristic Functions. We have
\[ \chi _{U\mathbin {\triangle }V}=\chi _{U}+\chi _{V}-2\chi _{U\cap V} \]
and thus, in particular, we have
\[ \chi _{U\mathbin {\triangle }V}\equiv \chi _{U}+\chi _{V}\ \ (\mathrm{mod}\ 2) \]
for each $U,V\in \mathcal{P}\webleft (X\webright )$.
-
Bijectivity. Given $U,V\in \mathcal{P}\webleft (X\webright )$, the maps
\begin{align*} U\mathbin {\triangle }- & \colon \mathcal{P}\webleft (X\webright ) \to \mathcal{P}\webleft (X\webright ),\\ -\mathbin {\triangle }V & \colon \mathcal{P}\webleft (X\webright ) \to \mathcal{P}\webleft (X\webright ) \end{align*}
are bijections with inverses given by
\begin{align*} \webleft (U\mathbin {\triangle }-\webright )^{-1} & = -\cup \webleft (U\cap -\webright ),\\ \webleft (-\mathbin {\triangle }V\webright )^{-1} & = -\cup \webleft (V\cap -\webright ). \end{align*}
Moreover, the map
is a bijection of $\mathcal{P}\webleft (X\webright )$ onto itself sending $U$ to $V$ and $V$ to $U$.
-
Interaction With Powersets and Groups. Let $X$ be a set.
-
The quadruple $\webleft (\mathcal{P}\webleft (X\webright ),\mathbin {\triangle },\text{Ø},\text{id}_{\mathcal{P}\webleft (X\webright )}\webright )$ is an abelian group.1
-
Every element of $\mathcal{P}\webleft (X\webright )$ has order $2$ with respect to $\mathbin {\triangle }$, and thus $\mathcal{P}\webleft (X\webright )$ is a Boolean group (i.e. an abelian $2$-group).
-
Interaction With Powersets and Vector Spaces I. The pair $\webleft (\mathcal{P}\webleft (X\webright ),\alpha _{\mathcal{P}\webleft (X\webright )}\webright )$ consisting of is an $\mathbb {F}_{2}$-vector space.
-
Interaction With Powersets and Vector Spaces II. If $X$ is finite, then:
- The set of singletons sets on the elements of $X$ forms a basis for the $\mathbb {F}_{2}$-vector space $\webleft (\mathcal{P}\webleft (X\webright ),\alpha _{\mathcal{P}\webleft (X\webright )}\webright )$ of Item 18.
- We have
\[ \dim \webleft (\mathcal{P}\webleft (X\webright )\webright )=\# {X}. \]
-
Interaction With Powersets and Rings. The quintuple $\webleft (\mathcal{P}\webleft (X\webright ),\mathbin {\triangle },\cap ,\text{Ø},X\webright )$ is a commutative ring.2
-
Interaction With Direct Images. We have a natural transformation
with components
\[ f_{*}\webleft (U\webright )\mathbin {\triangle }f_{*}\webleft (V\webright )\subset f_{*}\webleft (U\mathbin {\triangle }V\webright ) \]
indexed by $U,V\in \mathcal{P}\webleft (X\webright )$.
-
Interaction With Inverse Images. The diagram
i.e. we have
\[ f^{-1}\webleft (U\webright )\mathbin {\triangle }f^{-1}\webleft (V\webright )=f^{-1}\webleft (U\mathbin {\triangle }V\webright ) \]
for each $U,V\in \mathcal{P}\webleft (Y\webright )$.
-
Interaction With Direct Images With Compact Support. We have a natural transformation
with components
\[ f_{!}\webleft (U\mathbin {\triangle }V\webright )\subset f_{1}\webleft (U\webright )\mathbin {\triangle }f_{!}\webleft (V\webright ) \]
indexed by $U,V\in \mathcal{P}\webleft (X\webright )$.