Linear-Algebra-Done-Right · Unit 14 · Video 2 · Interactive Practice
Norms, Orthogonality and the Pythagorean Theorem
I Key Formulas
Key Insight: Work with ∥ v ∥ 2 = ⟨ v , v ⟩ \|v\|^2 = \langle v, v \rangle ∥ v ∥ 2 = ⟨ v , v ⟩ — both proofs expand an inner product instead of handling a square root.
II Scaling by a Complex Number
Multiplying z = ( 1 + 2 i , 2 , 4 i ) ∈ C 3 z = (1 + 2i, 2, 4i) \in \mathbf{C}^3 z = ( 1 + 2 i , 2 , 4 i ) ∈ C 3 by λ ∈ C \lambda \in \mathbf{C} λ ∈ C multiplies every coordinate's modulus, and the norm, by ∣ λ ∣ |\lambda| ∣ λ ∣ .
III Orthogonal Functions on [ − 1 , 1 ] [-1, 1] [ − 1 , 1 ]
With ⟨ f , g ⟩ = ∫ − 1 1 f g \langle f, g \rangle = \int_{-1}^{1} fg ⟨ f , g ⟩ = ∫ − 1 1 f g , orthogonality means the signed area under f g fg f g cancels to zero.
f
f(x) = 1
f(x) = x
f(x) = x²
IV The Pythagorean Theorem in R 2 \mathbf{R}^2 R 2
With the Euclidean inner product, the square on u + v u + v u + v equals the other two squares exactly when ⟨ u , v ⟩ = 0 \langle u, v \rangle = 0 ⟨ u , v ⟩ = 0 .
💡 Challenge: with u = ( 2 , 1 ) u = (2, 1) u = ( 2 , 1 ) fixed, find three different v v v that make the squares balance, and say what those v v v have in common.
V Quiz Questions
Problem 1 · A Norm in C 3 \mathbf{C}^3 C 3
Given: w = ( 2 − i , 1 + 3 i , 1 ) ∈ C 3 w = (2 - i,\ 1 + 3i,\ 1) \in \mathbf{C}^3 w = ( 2 − i , 1 + 3 i , 1 ) ∈ C 3 with the Euclidean inner product — find ∥ w ∥ \|w\| ∥ w ∥ .
a 6 \sqrt{6} 6
b 16 16 16
c 4 4 4
d 1 + 5 + 10 1 + \sqrt{5} + \sqrt{10} 1 + 5 + 10
✅ Correct! The squared moduli are
5 5 5 ,
10 10 10 and
1 1 1 , which add to
16 16 16 , so
∥ w ∥ = 4 \|w\| = 4 ∥ w ∥ = 4 .
❌ Not quite. That uses only the real parts
2 , 1 , 1 2, 1, 1 2 , 1 , 1 . Each term is
∣ w k ∣ 2 = ( Re w k ) 2 + ( Im w k ) 2 |w_k|^2 = (\operatorname{Re} w_k)^2 + (\operatorname{Im} w_k)^2 ∣ w k ∣ 2 = ( Re w k ) 2 + ( Im w k ) 2 , so the imaginary parts count too.
❌ Close, but… 16 16 16 is
∥ w ∥ 2 = ⟨ w , w ⟩ \|w\|^2 = \langle w, w \rangle ∥ w ∥ 2 = ⟨ w , w ⟩ . The norm is its square root.
❌ Not quite. The norm is not the sum of the moduli
∣ w k ∣ |w_k| ∣ w k ∣ — square the moduli, add, then take one square root.
❌ Not quite. Compute
∣ w 1 ∣ 2 + ∣ w 2 ∣ 2 + ∣ w 3 ∣ 2 |w_1|^2 + |w_2|^2 + |w_3|^2 ∣ w 1 ∣ 2 + ∣ w 2 ∣ 2 + ∣ w 3 ∣ 2 first.
Show solution
For the Euclidean inner product, ∥ w ∥ = ∣ w 1 ∣ 2 + ∣ w 2 ∣ 2 + ∣ w 3 ∣ 2 \|w\| = \sqrt{|w_1|^2 + |w_2|^2 + |w_3|^2} ∥ w ∥ = ∣ w 1 ∣ 2 + ∣ w 2 ∣ 2 + ∣ w 3 ∣ 2 . The squared moduli are
∣ 2 − i ∣ 2 = 4 + 1 = 5 , ∣ 1 + 3 i ∣ 2 = 1 + 9 = 10 , ∣ 1 ∣ 2 = 1. |2 - i|^2 = 4 + 1 = 5, \qquad |1 + 3i|^2 = 1 + 9 = 10, \qquad |1|^2 = 1. ∣2 − i ∣ 2 = 4 + 1 = 5 , ∣1 + 3 i ∣ 2 = 1 + 9 = 10 , ∣1 ∣ 2 = 1.
∥ w ∥ 2 = 5 + 10 + 1 = 16 , ∥ w ∥ = 4. \|w\|^2 = 5 + 10 + 1 = 16, \qquad \|w\| = 4. ∥ w ∥ 2 = 5 + 10 + 1 = 16 , ∥ w ∥ = 4.
The norm is 4 \mathbf{4} 4 .
Problem 2 · Pulling Out a Complex Scalar
Given: v = ( 1 , 2 i , 2 ) ∈ C 3 v = (1,\ 2i,\ 2) \in \mathbf{C}^3 v = ( 1 , 2 i , 2 ) ∈ C 3 with the Euclidean inner product and λ = − 3 + 4 i \lambda = -3 + 4i λ = − 3 + 4 i — find ∥ λ v ∥ \|\lambda v\| ∥ λ v ∥ .
a − 9 + 12 i -9 + 12i − 9 + 12 i
b 21 21 21
c 75 75 75
d 15 15 15
✅ Correct! ∥ λ v ∥ = ∣ λ ∣ ∥ v ∥ = 5 ⋅ 3 = 15 \|\lambda v\| = |\lambda|\,\|v\| = 5 \cdot 3 = 15 ∥ λ v ∥ = ∣ λ ∣ ∥ v ∥ = 5 ⋅ 3 = 15 .
❌ Not quite. A norm is a nonnegative real number. The scalar comes out as
∣ λ ∣ |\lambda| ∣ λ ∣ , not as
λ \lambda λ :
∥ λ v ∥ = ∣ λ ∣ ∥ v ∥ \|\lambda v\| = |\lambda|\,\|v\| ∥ λ v ∥ = ∣ λ ∣ ∥ v ∥ .
❌ Not quite. ∣ − 3 + 4 i ∣ |-3 + 4i| ∣ − 3 + 4 i ∣ is
( − 3 ) 2 + 4 2 = 5 \sqrt{(-3)^2 + 4^2} = 5 ( − 3 ) 2 + 4 2 = 5 , not
3 + 4 = 7 3 + 4 = 7 3 + 4 = 7 .
❌ Close, but… ∣ λ ∣ 2 = 25 |\lambda|^2 = 25 ∣ λ ∣ 2 = 25 belongs in
∥ λ v ∥ 2 = ∣ λ ∣ 2 ∥ v ∥ 2 \|\lambda v\|^2 = |\lambda|^2 \|v\|^2 ∥ λ v ∥ 2 = ∣ λ ∣ 2 ∥ v ∥ 2 . Taking square roots of both sides gives
∣ λ ∣ ∥ v ∥ |\lambda|\,\|v\| ∣ λ ∣ ∥ v ∥ .
❌ Not quite. Find
∥ v ∥ \|v\| ∥ v ∥ and
∣ λ ∣ |\lambda| ∣ λ ∣ separately.
Show solution
∥ v ∥ 2 = ∣ 1 ∣ 2 + ∣ 2 i ∣ 2 + ∣ 2 ∣ 2 = 1 + 4 + 4 = 9 , ∥ v ∥ = 3. \|v\|^2 = |1|^2 + |2i|^2 + |2|^2 = 1 + 4 + 4 = 9, \qquad \|v\| = 3. ∥ v ∥ 2 = ∣1 ∣ 2 + ∣2 i ∣ 2 + ∣2 ∣ 2 = 1 + 4 + 4 = 9 , ∥ v ∥ = 3.
∣ λ ∣ = ∣ − 3 + 4 i ∣ = 9 + 16 = 5. |\lambda| = |-3 + 4i| = \sqrt{9 + 16} = 5. ∣ λ ∣ = ∣ − 3 + 4 i ∣ = 9 + 16 = 5.
∥ λ v ∥ = ∣ λ ∣ ∥ v ∥ = 5 ⋅ 3 = 15. \|\lambda v\| = |\lambda|\,\|v\| = 5 \cdot 3 = 15. ∥ λ v ∥ = ∣ λ ∣ ∥ v ∥ = 5 ⋅ 3 = 15.
Direct check: λ v = ( − 3 + 4 i , − 8 − 6 i , − 6 + 8 i ) \lambda v = (-3 + 4i,\ -8 - 6i,\ -6 + 8i) λ v = ( − 3 + 4 i , − 8 − 6 i , − 6 + 8 i ) , whose squared moduli are 25 25 25 , 100 100 100 , 100 100 100 , summing to 225 = 15 2 225 = 15^2 225 = 1 5 2 .
Problem 3 · Orthogonal Polynomials on [ 0 , 1 ] [0, 1] [ 0 , 1 ]
Given: P 1 ( R ) \mathcal{P}_1(\mathbf{R}) P 1 ( R ) , the real polynomials of degree at most 1 1 1 , with the inner product ⟨ p , q ⟩ = ∫ 0 1 p q \langle p, q \rangle = \int_0^1 p q ⟨ p , q ⟩ = ∫ 0 1 pq . Let p ( x ) = x p(x) = x p ( x ) = x and q ( x ) = x + c q(x) = x + c q ( x ) = x + c — find the c ∈ R c \in \mathbf{R} c ∈ R that makes p p p and q q q orthogonal, then ∥ p + q ∥ 2 \|p + q\|^2 ∥ p + q ∥ 2 .
What is c c c ?
a c = − 1 2 c = -\tfrac{1}{2} c = − 2 1
b c = − 2 3 c = -\tfrac{2}{3} c = − 3 2
c c = 0 c = 0 c = 0
d c = − 3 2 c = -\tfrac{3}{2} c = − 2 3
What is ∥ p + q ∥ 2 \|p + q\|^2 ∥ p + q ∥ 2 ?
a 4 9 \tfrac{4}{9} 9 4
b 4 3 \tfrac{4}{3} 3 4
c 4 + 2 3 9 \tfrac{4 + 2\sqrt{3}}{9} 9 4 + 2 3
d 2 9 \tfrac{2}{9} 9 2
✅ Correct! ⟨ x , x − 2 3 ⟩ = 0 \langle x, x - \tfrac{2}{3} \rangle = 0 ⟨ x , x − 3 2 ⟩ = 0 , so
∥ p + q ∥ 2 = ∥ p ∥ 2 + ∥ q ∥ 2 = 1 3 + 1 9 = 4 9 \|p + q\|^2 = \|p\|^2 + \|q\|^2 = \tfrac{1}{3} + \tfrac{1}{9} = \tfrac{4}{9} ∥ p + q ∥ 2 = ∥ p ∥ 2 + ∥ q ∥ 2 = 3 1 + 9 1 = 9 4 .
❌ Not quite. That solves
∫ 0 1 ( x + c ) = 0 \int_0^1 (x + c) = 0 ∫ 0 1 ( x + c ) = 0 , which is
⟨ 1 , q ⟩ = 0 \langle 1, q \rangle = 0 ⟨ 1 , q ⟩ = 0 . The condition is
∫ 0 1 x ( x + c ) = 0 \int_0^1 x(x + c) = 0 ∫ 0 1 x ( x + c ) = 0 .
❌ Not quite. c = 0 c = 0 c = 0 makes
q = p q = p q = p , and
⟨ x , x ⟩ = ∫ 0 1 x 2 = 1 3 ≠ 0 \langle x, x \rangle = \int_0^1 x^2 = \tfrac{1}{3} \ne 0 ⟨ x , x ⟩ = ∫ 0 1 x 2 = 3 1 = 0 — a nonzero vector is never orthogonal to itself.
❌ Close, but… From
1 3 + c 2 = 0 \tfrac{1}{3} + \tfrac{c}{2} = 0 3 1 + 2 c = 0 you get
c = − 2 3 c = -\tfrac{2}{3} c = − 3 2 ; the fraction was inverted.
❌ Check c c c . Expand
∫ 0 1 ( x 2 + c x ) \int_0^1 (x^2 + c x) ∫ 0 1 ( x 2 + c x ) and set it to zero.
❌ Not quite. 4 3 = ∥ 2 x ∥ 2 \tfrac{4}{3} = \|2x\|^2 3 4 = ∥2 x ∥ 2 — that drops
c c c . Here
p + q = 2 x − 2 3 p + q = 2x - \tfrac{2}{3} p + q = 2 x − 3 2 .
❌ Not quite. That is
( ∥ p ∥ + ∥ q ∥ ) 2 (\|p\| + \|q\|)^2 ( ∥ p ∥ + ∥ q ∥ ) 2 . The Pythagorean theorem adds the
squares of the norms:
∥ p ∥ 2 + ∥ q ∥ 2 \|p\|^2 + \|q\|^2 ∥ p ∥ 2 + ∥ q ∥ 2 .
❌ Not quite. 2 9 = ∥ p ∥ 2 − ∥ q ∥ 2 = 1 3 − 1 9 \tfrac{2}{9} = \|p\|^2 - \|q\|^2 = \tfrac{1}{3} - \tfrac{1}{9} 9 2 = ∥ p ∥ 2 − ∥ q ∥ 2 = 3 1 − 9 1 . The Pythagorean theorem adds the squared norms.
❌ Check ∥ p + q ∥ 2 \|p + q\|^2 ∥ p + q ∥ 2 . Since
p ⊥ q p \perp q p ⊥ q , add
∥ p ∥ 2 = ∫ 0 1 x 2 \|p\|^2 = \int_0^1 x^2 ∥ p ∥ 2 = ∫ 0 1 x 2 and
∥ q ∥ 2 = ∫ 0 1 ( x − 2 3 ) 2 \|q\|^2 = \int_0^1 (x - \tfrac{2}{3})^2 ∥ q ∥ 2 = ∫ 0 1 ( x − 3 2 ) 2 .
Show solution
Step 1: Orthogonality.
⟨ p , q ⟩ = ∫ 0 1 x ( x + c ) d x = 1 3 + c 2 = 0 ⟹ c = − 2 3 . \langle p, q \rangle = \int_0^1 x(x + c)\,dx = \frac{1}{3} + \frac{c}{2} = 0 \;\Longrightarrow\; c = -\frac{2}{3}. ⟨ p , q ⟩ = ∫ 0 1 x ( x + c ) d x = 3 1 + 2 c = 0 ⟹ c = − 3 2 .
Step 2: The two norms squared.
∥ p ∥ 2 = ∫ 0 1 x 2 d x = 1 3 , ∥ q ∥ 2 = ∫ 0 1 ( x − 2 3 ) 2 d x = 1 3 − 2 3 + 4 9 = 1 9 . \|p\|^2 = \int_0^1 x^2\,dx = \frac{1}{3}, \qquad \|q\|^2 = \int_0^1 \Big(x - \frac{2}{3}\Big)^2 dx = \frac{1}{3} - \frac{2}{3} + \frac{4}{9} = \frac{1}{9}. ∥ p ∥ 2 = ∫ 0 1 x 2 d x = 3 1 , ∥ q ∥ 2 = ∫ 0 1 ( x − 3 2 ) 2 d x = 3 1 − 3 2 + 9 4 = 9 1 .
Step 3: Pythagorean theorem. ∥ p + q ∥ 2 = 1 3 + 1 9 = 4 9 \|p + q\|^2 = \frac{1}{3} + \frac{1}{9} = \frac{4}{9} ∥ p + q ∥ 2 = 3 1 + 9 1 = 9 4 .
Direct check: p + q = 2 x − 2 3 p + q = 2x - \tfrac{2}{3} p + q = 2 x − 3 2 and ∫ 0 1 ( 2 x − 2 3 ) 2 d x = 4 3 − 4 3 + 4 9 = 4 9 \int_0^1 (2x - \tfrac{2}{3})^2\,dx = \tfrac{4}{3} - \tfrac{4}{3} + \tfrac{4}{9} = \tfrac{4}{9} ∫ 0 1 ( 2 x − 3 2 ) 2 d x = 3 4 − 3 4 + 9 4 = 9 4 .
Problem 4 · Pythagoras in C 2 \mathbf{C}^2 C 2
Given: u = ( 1 , i ) u = (1, i) u = ( 1 , i ) and v = ( i , 1 ) v = (i, 1) v = ( i , 1 ) in C 2 \mathbf{C}^2 C 2 with the Euclidean inner product ⟨ w , z ⟩ = w 1 z 1 ‾ + w 2 z 2 ‾ \langle w, z \rangle = w_1 \overline{z_1} + w_2 \overline{z_2} ⟨ w , z ⟩ = w 1 z 1 + w 2 z 2 — find ⟨ u , v ⟩ \langle u, v \rangle ⟨ u , v ⟩ , then ∥ u + v ∥ \|u + v\| ∥ u + v ∥ .
What is ⟨ u , v ⟩ \langle u, v \rangle ⟨ u , v ⟩ ?
a 2 i 2i 2 i
b 2 2 2
c 0 0 0
d − 2 -2 − 2
What is ∥ u + v ∥ \|u + v\| ∥ u + v ∥ ?
a 4 4 4
b 2 2 2
c 2 2 2\sqrt{2} 2 2
d 2 \sqrt{2} 2
✅ Correct! u ⊥ v u \perp v u ⊥ v , so
∥ u + v ∥ 2 = 2 + 2 = 4 \|u + v\|^2 = 2 + 2 = 4 ∥ u + v ∥ 2 = 2 + 2 = 4 and
∥ u + v ∥ = 2 \|u + v\| = 2 ∥ u + v ∥ = 2 .
❌ Not quite. 1 ⋅ i + i ⋅ 1 = 2 i 1 \cdot i + i \cdot 1 = 2i 1 ⋅ i + i ⋅ 1 = 2 i forgets the conjugate on the second slot:
⟨ u , v ⟩ = 1 ⋅ i ‾ + i ⋅ 1 ‾ = − i + i \langle u, v \rangle = 1 \cdot \overline{i} + i \cdot \overline{1} = -i + i ⟨ u , v ⟩ = 1 ⋅ i + i ⋅ 1 = − i + i .
❌ Not quite. 2 = ∣ 1 ∣ ∣ i ∣ + ∣ i ∣ ∣ 1 ∣ 2 = |1|\,|i| + |i|\,|1| 2 = ∣1∣ ∣ i ∣ + ∣ i ∣ ∣1∣ uses moduli. The inner product keeps the complex numbers:
1 ⋅ i ‾ + i ⋅ 1 ‾ = − i + i 1 \cdot \overline{i} + i \cdot \overline{1} = -i + i 1 ⋅ i + i ⋅ 1 = − i + i .
❌ Not quite. The inner product pairs each coordinate of
u u u with the conjugate of the same coordinate of
v v v :
1 ⋅ i ‾ + i ⋅ 1 ‾ = − i + i 1 \cdot \overline{i} + i \cdot \overline{1} = -i + i 1 ⋅ i + i ⋅ 1 = − i + i .
❌ Check ⟨ u , v ⟩ \langle u, v \rangle ⟨ u , v ⟩ . Conjugate the coordinates of
v v v :
i ‾ = − i \overline{i} = -i i = − i ,
1 ‾ = 1 \overline{1} = 1 1 = 1 .
❌ Close, but… 4 4 4 is
∥ u + v ∥ 2 \|u + v\|^2 ∥ u + v ∥ 2 . Take the square root.
❌ Not quite. 2 2 = ∥ u ∥ + ∥ v ∥ 2\sqrt{2} = \|u\| + \|v\| 2 2 = ∥ u ∥ + ∥ v ∥ . The Pythagorean theorem adds squares:
∥ u + v ∥ 2 = ∥ u ∥ 2 + ∥ v ∥ 2 \|u + v\|^2 = \|u\|^2 + \|v\|^2 ∥ u + v ∥ 2 = ∥ u ∥ 2 + ∥ v ∥ 2 .
❌ Not quite. 2 \sqrt{2} 2 is
∥ u ∥ \|u\| ∥ u ∥ alone. Both legs count:
∥ u + v ∥ 2 = ∥ u ∥ 2 + ∥ v ∥ 2 = 2 + 2 \|u + v\|^2 = \|u\|^2 + \|v\|^2 = 2 + 2 ∥ u + v ∥ 2 = ∥ u ∥ 2 + ∥ v ∥ 2 = 2 + 2 .
❌ Check ∥ u + v ∥ \|u + v\| ∥ u + v ∥ . Each of
u u u ,
v v v has norm squared
∣ 1 ∣ 2 + ∣ i ∣ 2 = 2 |1|^2 + |i|^2 = 2 ∣1 ∣ 2 + ∣ i ∣ 2 = 2 .
Show solution
Step 1: Inner product.
⟨ u , v ⟩ = 1 ⋅ i ‾ + i ⋅ 1 ‾ = − i + i = 0 , \langle u, v \rangle = 1 \cdot \overline{i} + i \cdot \overline{1} = -i + i = 0, ⟨ u , v ⟩ = 1 ⋅ i + i ⋅ 1 = − i + i = 0 ,
so u u u and v v v are orthogonal.
Step 2: Pythagorean theorem. ∥ u ∥ 2 = ∣ 1 ∣ 2 + ∣ i ∣ 2 = 2 \|u\|^2 = |1|^2 + |i|^2 = 2 ∥ u ∥ 2 = ∣1 ∣ 2 + ∣ i ∣ 2 = 2 and ∥ v ∥ 2 = ∣ i ∣ 2 + ∣ 1 ∣ 2 = 2 \|v\|^2 = |i|^2 + |1|^2 = 2 ∥ v ∥ 2 = ∣ i ∣ 2 + ∣1 ∣ 2 = 2 , so
∥ u + v ∥ 2 = ∥ u ∥ 2 + ∥ v ∥ 2 = 4 , ∥ u + v ∥ = 2. \|u + v\|^2 = \|u\|^2 + \|v\|^2 = 4, \qquad \|u + v\| = 2. ∥ u + v ∥ 2 = ∥ u ∥ 2 + ∥ v ∥ 2 = 4 , ∥ u + v ∥ = 2.
Direct check: u + v = ( 1 + i , 1 + i ) u + v = (1 + i,\ 1 + i) u + v = ( 1 + i , 1 + i ) and ∣ 1 + i ∣ 2 + ∣ 1 + i ∣ 2 = 2 + 2 = 4 |1 + i|^2 + |1 + i|^2 = 2 + 2 = 4 ∣1 + i ∣ 2 + ∣1 + i ∣ 2 = 2 + 2 = 4 .
Solved: 0 / 4