Mathematical-Reasoning · Unit 21 · Video 1 · Interactive Practice

Polygon Diagonals and Derivatives: A Formula Is Not a Function

IKey Formulas

FormulaNameMeaning
D=N−{1,2}D = \mathbb{N} - \{1, 2\},   d ⁣:D→N∪{0}\;d\colon D \to \mathbb{N} \cup \{0\}Diagonal countd(n)d(n) is the number of diagonals of a convex polygon with nn sides
d(n)=n(n−3)2d(n) = \dfrac{n(n-3)}{2}Counting the diagonalsEach of the nn vertices starts n−3n - 3 diagonals, and each diagonal is counted once at each of its two ends
f ⁣:R→Rf\colon \mathbb{R} \to \mathbb{R},   f(x)=x(x−3)2\;f(x) = \dfrac{x(x-3)}{2}Same expression, new functiond(n)=f(n)d(n) = f(n) for every n∈Dn \in D, yet the domains and codomains differ, so d≠fd \neq f
f↦f′f \mapsto f'Differentiation as a functionInputs: differentiable real functions; outputs: real functions. ∣x∣\lvert x \rvert is not an input

Key Insight: A function is its domain, its codomain and its rule together. A formula that matches a function's outputs does not settle the domain or the codomain, so it does not by itself make two functions the same.

IIVisualization 1 — Counting Diagonals Vertex by Vertex

Each vertex starts n−3n - 3 diagonals and every diagonal has two ends, so d(n)=n(n−3)2d(n) = \frac{n(n-3)}{2}.

IIIVisualization 2 — The Formula Beyond the Polygons

The formula f(x)=x(x−3)2f(x) = \frac{x(x-3)}{2} accepts every real xx; the count dd accepts only n=3,4,5,…n = 3, 4, 5, \dots

IVVisualization 3 — Differentiation as a Function

Differentiation sends each differentiable input ff to exactly one output f′f'; different inputs may share an output.

VQuiz Questions

Problem 1 · Handshakes Versus a Formula

Given: for n∈Nn \in \mathbb{N}, let h(n)h(n) be the number of handshakes when each pair of nn people shakes hands exactly once, so h ⁣:N→N∪{0}h\colon \mathbb{N} \to \mathbb{N} \cup \{0\}. Each person shakes n−1n - 1 hands and each handshake involves two people, so h(n)=n(n−1)2h(n) = \frac{n(n-1)}{2}. Let g ⁣:R→Rg\colon \mathbb{R} \to \mathbb{R} be g(x)=x(x−1)2g(x) = \frac{x(x-1)}{2}. Which statement is true?

✅ Correct! The rules give the same number at every n∈Nn \in \mathbb{N}, but hh accepts only natural numbers while gg accepts every real number, and the codomains N∪{0}\mathbb{N} \cup \{0\} and R\mathbb{R} differ too.
❌ Not quite. Agreeing on N\mathbb{N} is not enough: gg also accepts inputs such as 12\tfrac{1}{2}, with g(12)=−18g\left(\tfrac{1}{2}\right) = -\tfrac{1}{8}, and hh has no output there at all.
❌ Not quite. The two rules never disagree on N\mathbb{N}: the handshake count really is n(n−1)2\frac{n(n-1)}{2}, for instance h(1)=0=g(1)h(1) = 0 = g(1) and h(4)=6=g(4)h(4) = 6 = g(4). The difference lies elsewhere.
❌ Not quite. Changing the codomain leaves the domains unequal: hh would still be defined only on N\mathbb{N}, while gg is defined on R\mathbb{R}.
Show solution

Compare the three parts of each function.

  • Rule: h(n)h(n) counts handshakes. Each of the nn people shakes n−1n - 1 hands, which gives n(n−1)n(n-1) handshake ends, and each handshake has two ends, so h(n)=n(n−1)2=g(n)h(n) = \frac{n(n-1)}{2} = g(n) for every n∈Nn \in \mathbb{N}.
  • Domain: N\mathbb{N} for hh, but R\mathbb{R} for gg. For example g(−2)=(−2)(−3)2=3g(-2) = \frac{(-2)(-3)}{2} = 3, while h(−2)h(-2) is not defined.
  • Codomain: N∪{0}\mathbb{N} \cup \{0\} for hh, but R\mathbb{R} for gg; g(12)=−18g\left(\tfrac{1}{2}\right) = -\tfrac{1}{8} is not a count of anything.

Matching outputs on N\mathbb{N} do not make the functions equal, because the domains and codomains differ. So h≠gh \neq g, exactly as d≠fd \neq f in the video.

Problem 2 · Triangles from Points on a Circle

Given: for n∈Nn \in \mathbb{N}, let t(n)t(n) be the number of triangles whose three vertices are chosen from nn points marked on a circle, so t ⁣:N→N∪{0}t\colon \mathbb{N} \to \mathbb{N} \cup \{0\}. For n≥3n \ge 3, counting gives t(n)=n(n−1)(n−2)6t(n) = \frac{n(n-1)(n-2)}{6}. Let p ⁣:R→Rp\colon \mathbb{R} \to \mathbb{R} be p(x)=x(x−1)(x−2)6p(x) = \frac{x(x-1)(x-2)}{6}.

What is p(−1)p(-1)?

For which inputs xx is t(x)=p(x)t(x) = p(x) a true statement?

✅ Correct! p(−1)=−1p(-1) = -1 is a value of pp at an input tt does not accept, and −1∉N∪{0}-1 \notin \mathbb{N} \cup \{0\} anyway. The two functions agree on all of N\mathbb{N}, including n=1n = 1 and n=2n = 2, where both give 00.
❌ Check the sign. The product (−1)(−2)(−3)(-1)(-2)(-3) has three negative factors, so it is −6-6, not 66.
❌ Not quite. "Fewer than three points, so no triangles" is a thought about tt, and tt does not even accept −1-1. The question asks about pp, a formula on R\mathbb{R}: substitute x=−1x = -1 into x(x−1)(x−2)6\frac{x(x-1)(x-2)}{6}.
❌ Almost. The product (−1)(−2)(−3)=−6(-1)(-2)(-3) = -6 is right, but the formula still divides it by 66.
❌ Not quite. Check n=1n = 1 and n=2n = 2: with fewer than three points there are no triangles, so t(1)=t(2)=0t(1) = t(2) = 0, and p(1)=p(2)=0p(1) = p(2) = 0 as well.
❌ Not quite. tt accepts only natural numbers, so t(32)t\left(\tfrac{3}{2}\right) is not defined even though p(32)=−116p\left(\tfrac{3}{2}\right) = -\tfrac{1}{16} is.
❌ Not quite. At x=−1x = -1 the formula gives p(−1)=−1p(-1) = -1, but t(−1)t(-1) does not exist: no one marks −1-1 points on a circle.
❌ Not quite. Substitute x=−1x = -1 into each factor and multiply.
❌ Not quite. The statement needs both t(x)t(x) and p(x)p(x) to exist and be equal.
Show solution

The value:

p(−1)=(−1)(−1−1)(−1−2)6=(−1)(−2)(−3)6=−66=−1p(-1) = \frac{(-1)(-1-1)(-1-2)}{6} = \frac{(-1)(-2)(-3)}{6} = \frac{-6}{6} = -1

The input −1-1 is not in the domain N\mathbb{N} of tt, and the output −1-1 is not in its codomain N∪{0}\mathbb{N} \cup \{0\}.

Where they agree: t(x)t(x) exists only for x∈Nx \in \mathbb{N}. For n≥3n \ge 3 the stem gives t(n)=p(n)t(n) = p(n). For the two remaining natural numbers, one or two points form no triangle, so t(1)=t(2)=0t(1) = t(2) = 0, and

p(1)=1⋅0⋅(−1)6=0,p(2)=2⋅1⋅06=0.p(1) = \frac{1 \cdot 0 \cdot (-1)}{6} = 0, \qquad p(2) = \frac{2 \cdot 1 \cdot 0}{6} = 0.

So t(n)=p(n)t(n) = p(n) for every natural number nn. Even so, t≠pt \neq p: the domains N\mathbb{N} and R\mathbb{R} differ.

Problem 3 · Which Function Is an Input?

Given: let FD\mathcal{F}_D be the set of functions from R\mathbb{R} to R\mathbb{R} that have a derivative at every real number; it is the domain of differentiation in this problem. Which function is an element of FD\mathcal{F}_D?

✅ Correct! s(x)=x2s(x) = x^2 for x≥0x \ge 0 and s(x)=−x2s(x) = -x^2 for x<0x \lt 0; both pieces have slope 00 at 00, so ss has a derivative everywhere, with s′(x)=2∣x∣s'(x) = 2\lvert x \rvert.
❌ Not quite. uu is defined on all of R\mathbb{R}, but its graph has a vertical tangent at 00: u(h)−u(0)h=h−2/3\frac{u(h) - u(0)}{h} = h^{-2/3} grows without bound as h→0h \to 0, so u′(0)u'(0) does not exist.
❌ Not quite. ww is the absolute value graph moved to the right by 22: it has a corner at x=2x = 2, with slope −1-1 from the left and 11 from the right.
❌ Not quite. The pieces meet at (1,1)(1, 1), so there is no gap, but the slope from the left is 2⋅1=22 \cdot 1 = 2 and the slope from the right is 11. No single slope fits both sides at x=1x = 1.
Show solution

An element of FD\mathcal{F}_D must be defined on all of R\mathbb{R} and have a derivative at every real number. Each wrong option fails at one point.

  • u(x)=x3u(x) = \sqrt[3]{x}: for x≠0x \neq 0, u′(x)=13x−2/3u'(x) = \frac{1}{3}x^{-2/3}, but at 00 the difference quotient h3h=h−2/3\frac{\sqrt[3]{h}}{h} = h^{-2/3} has no finite limit. Not an input.
  • w(x)=∣x−2∣w(x) = \lvert x - 2 \rvert: slope −1-1 for x<2x \lt 2 and 11 for x>2x \gt 2; no derivative at 22. Not an input.
  • qq: slope from the left at 11 is ddxx2=2x=2\frac{d}{dx}x^2 = 2x = 2, slope from the right is 11; no derivative at 11. Not an input.

For s(x)=x∣x∣s(x) = x\lvert x \rvert: if x>0x \gt 0 then s(x)=x2s(x) = x^2 and s′(x)=2xs'(x) = 2x; if x<0x \lt 0 then s(x)=−x2s(x) = -x^2 and s′(x)=−2xs'(x) = -2x. At 00,

s(h)−s(0)h=h∣h∣h=∣h∣→0,\frac{s(h) - s(0)}{h} = \frac{h\lvert h \rvert}{h} = \lvert h \rvert \to 0,

so s′(0)=0s'(0) = 0. Thus s∈FDs \in \mathcal{F}_D, and differentiation sends it to the single output s′(x)=2∣x∣s'(x) = 2\lvert x \rvert. That output is itself not in FD\mathcal{F}_D, but it is a real function, so it lies in the codomain.

Problem 4 · Inputs and Outputs of Differentiation

Given: FD\mathcal{F}_D as in Problem 3, F\mathcal{F} the set of all functions from R\mathbb{R} to R\mathbb{R}, and T ⁣:FD→FT\colon \mathcal{F}_D \to \mathcal{F} defined by T(f)=f′T(f) = f'.

For m(x)=x2e3xm(x) = x^2 e^{3x}, what is T(m)(x)T(m)(x)?

For f(x)=x3+4xf(x) = x^3 + 4x and g(x)=x3+4x+10g(x) = x^3 + 4x + 10, which statement is true?

✅ Correct! T(m)(x)=2xe3x+3x2e3xT(m)(x) = 2xe^{3x} + 3x^2e^{3x}, and ff and gg are two different inputs sent to the same output 3x2+43x^2 + 4, which a function allows.
❌ Not quite. That differentiates only the factor x2x^2. The product rule also needs x2x^2 times the derivative of e3xe^{3x}.
❌ Not quite. The derivative of a product is not the product of the derivatives 2x⋅3e3x2x \cdot 3e^{3x}; use the product rule.
❌ Close. The chain rule gives ddxe3x=3e3x\frac{d}{dx}e^{3x} = 3e^{3x}, not e3xe^{3x}.
❌ Not quite. The constant 1010 has derivative 00, so T(g)(x)=3x2+4=T(f)(x)T(g)(x) = 3x^2 + 4 = T(f)(x).
❌ Not quite. A function must give each input exactly one output. Nothing forbids two different inputs from sharing an output; each of ff and gg still has exactly one derivative.
❌ Not quite. f(0)=0f(0) = 0 but g(0)=10g(0) = 10, so f≠gf \neq g. Equal outputs tell us nothing about whether the inputs were equal.
❌ Not quite. Use the product rule on x2⋅e3xx^2 \cdot e^{3x}.
❌ Not quite. Differentiate both ff and gg before deciding.
Show solution

Part 1. The product rule, with the chain rule for e3xe^{3x}:

T(m)(x)=ddx(x2)e3x+x2ddx(e3x)=2xe3x+3x2e3x.T(m)(x) = \frac{d}{dx}\left(x^2\right) e^{3x} + x^2 \frac{d}{dx}\left(e^{3x}\right) = 2x e^{3x} + 3x^2 e^{3x}.

Part 2. By the power rule term by term, and since a constant has derivative 00,

T(f)(x)=3x2+4,T(g)(x)=3x2+4+0=3x2+4.T(f)(x) = 3x^2 + 4, \qquad T(g)(x) = 3x^2 + 4 + 0 = 3x^2 + 4.

So T(f)=T(g)T(f) = T(g) even though f≠gf \neq g (for instance f(0)=0f(0) = 0 and g(0)=10g(0) = 10). This does not conflict with TT being a function: the requirement is that each input ff in FD\mathcal{F}_D has exactly one output, and each differentiable function has exactly one derivative. Two inputs sharing one output is allowed.

Solved: 0 / 4