If
[math]|A| \href{/cs2800/wiki/index.php/%E2%89%A4}{≤} |B|
[/math] and
[math]|B| \href{/cs2800/wiki/index.php/%E2%89%A4}{≤} |C|
[/math] then
[math]|A| \href{/cs2800/wiki/index.php/%E2%89%A4}{≤} |C|
[/math].
In other words, the
[math]\href{/cs2800/wiki/index.php/%E2%89%A4}{≤}
[/math] relation on
cardinalities is
transitive.
Proof: (using definitions)
Suppose
[math]|A| \href{/cs2800/wiki/index.php/%E2%89%A4}{≤} |B|
[/math] and
[math]|B| \href{/cs2800/wiki/index.php/%E2%89%A4}{≤} |C|
[/math]. Then
by definition,
there exists injections [math]f : A \href{/cs2800/wiki/index.php/%E2%86%92}{→} B
[/math] and
[math]g : B \href{/cs2800/wiki/index.php/%E2%86%92}{→} C
[/math].
I claim [math]g \href{/cs2800/wiki/index.php/%5Ccirc}{\circ} f : A \href{/cs2800/wiki/index.php/%E2%86%92}{→} C
[/math] is also an injection. To see this, suppose that [math]g \href{/cs2800/wiki/index.php/%5Ccirc}{\circ} f(x_1) = g \href{/cs2800/wiki/index.php/%5Ccirc}{\circ} f(x_2)
[/math]. We want to show that [math]x_1 = x_2
[/math].
Since [math]g
[/math] is
injective, we have
[math]f(x_1) = f(x_2)
[/math]. Since
[math]f
[/math] is also
injective, we have
[math]x_1 = x_2
[/math], as desired.
We can use inverses to give an alternate proof:
Proof: (using inverses)
Suppose
[math]|A| \href{/cs2800/wiki/index.php/%E2%89%A4}{≤} |B|
[/math] and
[math]|B| \href{/cs2800/wiki/index.php/%E2%89%A4}{≤} |C|
[/math]. Then
by definition,
there exists injections [math]f : A \href{/cs2800/wiki/index.php/%E2%86%92}{→} B
[/math] and
[math]g : B \href{/cs2800/wiki/index.php/%E2%86%92}{→} C
[/math].
I claim [math]g \href{/cs2800/wiki/index.php/%5Ccirc}{\circ} f : A \href{/cs2800/wiki/index.php/%E2%86%92}{→} C
[/math] is also an injection. To see this, note that [math]f
[/math] and [math]g
[/math] have left inverses [math]f' : B \href{/cs2800/wiki/index.php/%E2%86%92}{→} A
[/math] and [math]g' : C \href{/cs2800/wiki/index.php/%E2%86%92}{→} B
[/math].
Now, [math]f' \href{/cs2800/wiki/index.php/%5Ccirc}{\circ} g'
[/math] is a left inverse of [math]g \href{/cs2800/wiki/index.php/%5Ccirc}{\circ} f
[/math], because for any [math]x \href{/cs2800/wiki/index.php/%E2%88%88}{∈} A
[/math],
[math]\begin{align}
(f' \circ g') \circ (g \circ f)(x)
&= (f' \href{/cs2800/wiki/index.php/%5Ccirc}{\circ} g')(g(f(x))) && \href{/cs2800/wiki/index.php/%5Ccirc}{\text{by definition}} \\
&= f'(g'(g(f(x)))) && \href{/cs2800/wiki/index.php/%5Ccirc}{\text{by definition}} \\
&= f'(f(x)) && \href{/cs2800/wiki/index.php/Left_inverse}{\text{since g' is a left inverse of g; g'(g(y)) = y}} \\
&= x && \text{since f' is a left inverse of f}
\end{align}
[/math]
as required.
Since [math]g \href{/cs2800/wiki/index.php/%5Ccirc}{\circ} f
[/math] has a
left inverse, it is
injective, so by
definition, we have
[math]|A| \href{/cs2800/wiki/index.php/%E2%89%A4}{≤} |C|
[/math] as required.