Linear AlgebraLinear Transformations7 of 7
Linear Transformations · 7

Similar Matrices: Same Map, New Basis

A linear transformation is a physical action on space, like a stretch or a rotation, and the matrix we write for it depends on the coordinate grid we lay down to measure it.

1

Same Map, Different Grids

Two matrices are similar when they describe the exact same linear map in two different coordinate bases.

Before any grid is drawn, stretching a flat sheet by a factor of 33 along the 45∘45^\circ line while leaving the perpendicular direction unchanged is a single geometric action. Laying down a coordinate grid only decides how we record where the basis vectors land.

Same Map, Different Grids
tilted grid\text{tilted grid}
standard grid\text{standard grid}
b⃗1\htmlClass{lc-green}{\vec{b}_1}
3b⃗1\htmlClass{lc-green}{3\vec{b}_1}
b⃗2\htmlClass{lc-pink}{\vec{b}_2}
ı^\hat{\imath}
ȷ^\hat{\jmath}
[21]\begin{bmatrix} 2 \\ 1 \end{bmatrix}
[12]\begin{bmatrix} 1 \\ 2 \end{bmatrix}

Recording that single 45∘45^\circ stretch in our standard grid versus a friend's tilted grid (b⃗1\bone{\vec{b}_1} along the 45∘45^\circ stretch, b⃗2\btwo{\vec{b}_2} perpendicular to it) gives two different matrices for the same map:

A=[2112]⏟Standard grid (ı^,ȷ^)andB=[3001]⏟Tilted grid (b⃗1,b⃗2)\displaystyle \underbrace{\themap{A} = \begin{bmatrix} 2 & 1 \\ 1 & 2 \end{bmatrix}}_{\substack{\text{Standard grid } (\hat{\imath}, \hat{\jmath})}} \qquad \text{and} \qquad \underbrace{\themap{B} = \begin{bmatrix} \bone{3} & \btwo{0} \\ \bone{0} & \btwo{1} \end{bmatrix}}_{\substack{\text{Tilted grid } (\bone{\vec{b}_1}, \btwo{\vec{b}_2})}}

As the coordinate grid turns over the stretched sheet, all four entries of the matrix change while the 45∘45^\circ oval stays fixed, and the off-diagonal entries drop to 00 right when the basis vectors line up with the stretch and its perpendicular direction.

Turning the coordinate grid changes the matrix for the exact same 45° stretch, and the off-diagonal numbers drop to 0 when the grid lines up with the stretch.
2

The Three-Step Detour: B=P−1APB = P^{-1}AP

Suppose we know the matrix A=[2112]\themap{A} = \begin{bmatrix} 2 & 1 \\ 1 & 2 \end{bmatrix} for a map in our standard grid, and we want to find the matrix B\themap{B} that describes that same map in the tilted basis b⃗1=[11]\bone{\vec{b}_1} = \bone{\begin{bmatrix} 1 \\ 1 \end{bmatrix}}, b⃗2=[−11]\btwo{\vec{b}_2} = \btwo{\begin{bmatrix} -1 \\ 1 \end{bmatrix}}.

To connect the two grids, we put the new basis vectors into the columns of a change-of-basis matrix PP:

P=[b⃗1b⃗2]=[1−111]\displaystyle P = \begin{bmatrix} \bone{\vec{b}_1} & \btwo{\vec{b}_2} \end{bmatrix} = \begin{bmatrix} \bone{1} & \btwo{-1} \\ \bone{1} & \btwo{1} \end{bmatrix}

Multiplying by PP takes coordinates from their tilted grid and gives the vector in our standard grid, while P−1P^{-1} translates from our standard grid back into their grid:

P⏟Their grid → Our grid and P−1⏟Our grid → Their grid\displaystyle \underbrace{P}_{\text{Their grid } \to \text{ Our grid}} \text{ and } \underbrace{P^{-1}}_{\text{Our grid } \to \text{ Their grid}}

Because A\themap{A} only accepts standard coordinates, we cannot multiply A\themap{A} directly by a vector written in their grid. Instead, we translate the vector into our grid, apply A\themap{A}, and translate the answer back.

The Three-Step Detour: B = P⁻¹AP
b⃗1\htmlClass{lc-green}{\vec{b}_1}
b⃗2\htmlClass{lc-pink}{\vec{b}_2}
A=[2112]⟷?B=[3001]\htmlClass{lc-violet}{A} = \begin{bmatrix} 2 & 1 \\ 1 & 2 \end{bmatrix} \quad \overset{\textstyle ?}{\longleftrightarrow} \quad \htmlClass{lc-violet}{B} = \begin{bmatrix} \htmlClass{lc-green}{3} & \htmlClass{lc-pink}{0} \\ \htmlClass{lc-green}{0} & \htmlClass{lc-pink}{1} \end{bmatrix}

Reading from right to left in the order the matrices multiply an input vector x⃗\arrowx{\vec{x}} in their grid gives the formula for the similar matrix B\themap{B}:

B=P−1⏟3. Back totheir gridA⏟2. Map inour gridP⏟1. To ourgrid\displaystyle \begin{aligned} \themap{B} &= \underbrace{P^{-1}}_{\substack{\text{3. Back to} \\ \text{their grid}}} \quad \underbrace{\themap{A}}_{\substack{\text{2. Map in} \\ \text{our grid}}} \quad \underbrace{P}_{\substack{\text{1. To our} \\ \text{grid}}} \end{aligned}

Text solution:

When the stretched vectors Ab⃗1\themap{A}\bone{\vec{b}_1} and Ab⃗2\themap{A}\btwo{\vec{b}_2} are simple multiples of b⃗1\bone{\vec{b}_1} and b⃗2\btwo{\vec{b}_2}, you can see their new coordinates right away in Step 3 without writing out P−1P^{-1}.

For every input vector x⃗\arrowx{\vec{x}} in the tilted grid, translating to the standard grid with PP, stretching with A\themap{A}, and translating back with P−1P^{-1} lands on the exact same arrow and tilted coordinates as multiplying by B\themap{B} directly.

Translating any arrow to our standard grid, stretching it there, and translating it back always gives the exact same arrow and tilted coordinates as stretching it on the tilted grid directly.
3

Changing to a Skewed Basis

In the tilted basis above, the basis vectors lined up with the stretch, so we could see the new coordinates in Step 3 right away. Now we want the matrix B\themap{B} for the same stretch A=[2112]\themap{A} = \begin{bmatrix} 2 & 1 \\ 1 & 2 \end{bmatrix} in a skewed basis u⃗1=[23]\bone{\vec{u}_1} = \bone{\begin{bmatrix} 2 \\ 3 \end{bmatrix}}, u⃗2=[12]\btwo{\vec{u}_2} = \btwo{\begin{bmatrix} 1 \\ 2 \end{bmatrix}} that does not line up with the stretch at all.

When you cannot tell by looking how many u⃗1\bone{\vec{u}_1}'s and u⃗2\btwo{\vec{u}_2}'s it takes to build the stretched vectors, multiplying on the left by P−1P^{-1} does Step 3 for you.

Changing to a Skewed Basis
u⃗1\htmlClass{lc-green}{\vec{u}_1}
u⃗2\htmlClass{lc-pink}{\vec{u}_2}
A=[2112]B=  ?\begin{aligned} \htmlClass{lc-violet}{A} &= \begin{bmatrix} 2 & 1 \\ 1 & 2 \end{bmatrix} \\[0.5em] \htmlClass{lc-violet}{B} &= \;? \end{aligned}

Text solution:

So in the skewed basis u⃗1,u⃗2\bone{\vec{u}_1}, \btwo{\vec{u}_2}, the same 45∘45^\circ stretch is described by B=[63−5−2]\themap{B} = \begin{bmatrix} \bone{6} & \btwo{3} \\ \bone{-5} & \btwo{-2} \end{bmatrix}, where Column 1 tells us Au⃗1=6u⃗1−5u⃗2\themap{A}\bone{\vec{u}_1} = 6\bone{\vec{u}_1} - 5\btwo{\vec{u}_2} and Column 2 tells us Au⃗2=3u⃗1−2u⃗2\themap{A}\btwo{\vec{u}_2} = 3\bone{\vec{u}_1} - 2\btwo{\vec{u}_2}.

4

Back to the Standard Grid: A=PBP−1A = PBP^{-1}

Now suppose we start from the geometry of the stretch along b⃗1=[11]\bone{\vec{b}_1} = \bone{\begin{bmatrix} 1 \\ 1 \end{bmatrix}} and b⃗2=[−11]\btwo{\vec{b}_2} = \btwo{\begin{bmatrix} -1 \\ 1 \end{bmatrix}}, so we already know the easy matrix B=[3001]\themap{B} = \begin{bmatrix} \bone{3} & \btwo{0} \\ \bone{0} & \btwo{1} \end{bmatrix} in their tilted grid and want to find A\themap{A} in our standard grid.

To transform a vector that starts in our standard grid, we take the same three-step detour in reverse: translate from our grid to their tilted grid, apply the easy matrix B\themap{B}, and translate back to our grid.

Back to the Standard Grid: A = PBP⁻¹
b⃗1\htmlClass{lc-green}{\vec{b}_1}
b⃗2\htmlClass{lc-pink}{\vec{b}_2}
ı^\hat{\imath}
ȷ^\hat{\jmath}
B=[3001]A=  ?\begin{aligned} \htmlClass{lc-violet}{B} &= \begin{bmatrix} \htmlClass{lc-green}{3} & \htmlClass{lc-pink}{0} \\ \htmlClass{lc-green}{0} & \htmlClass{lc-pink}{1} \end{bmatrix} \\[0.5em] \htmlClass{lc-violet}{A} &= \;? \end{aligned}

Reading from right to left for an input vector starting in our standard grid gives A\themap{A}:

A=P⏟3. Back toour gridB⏟2. Map intheir gridP−1⏟1. To theirgrid\displaystyle \begin{aligned} \themap{A} &= \underbrace{P}_{\substack{\text{3. Back to} \\ \text{our grid}}} \quad \underbrace{\themap{B}}_{\substack{\text{2. Map in} \\ \text{their grid}}} \quad \underbrace{P^{-1}}_{\substack{\text{1. To their} \\ \text{grid}}} \end{aligned}

Text solution:

This detour also gives a fast way to compute high powers of A\themap{A}: rather than multiplying the harder matrix A\themap{A} ten times in the standard grid, translate to the tilted basis once, apply the diagonal matrix B\themap{B} ten times, and translate back at the end:

A10\displaystyle \themap{A^{10}} =PB10P−1\displaystyle = P\themap{B^{10}}P^{-1}
5

Testing Whether Matrices Are Similar

Given two matrices, we want to find out whether they are similar, meaning they can represent the same map in two different bases, without solving for a basis matrix PP.

Testing Whether Matrices Are Similar
A=[2112]C=[3112]\displaystyle {\themap{A} = \begin{bmatrix} 2 & 1 \\ 1 & 2 \end{bmatrix} \quad \themap{C} = \begin{bmatrix} 3 & 1 \\ 1 & 2 \end{bmatrix}}
similar?\displaystyle \text{similar?}

Because similar matrices represent the exact same geometric map, every property that belongs to the map itself must match across all of them:

I=[1001]\htmlClass{lc-violet}{I} = \begin{bmatrix} 1 & 0 \\ 0 & 1 \end{bmatrix}
S=[1101]\htmlClass{lc-violet}{S} = \begin{bmatrix} 1 & 1 \\ 0 & 1 \end{bmatrix}
ı^\hat{\imath}
ȷ^\hat{\jmath}
ı^\hat{\imath}
ȷ^\hat{\jmath}
[11]\begin{bmatrix} 1 \\ 1 \end{bmatrix}

Text solution:

Across every skewed basis u⃗1,u⃗2\bone{\vec{u}_1}, \btwo{\vec{u}_2} for the stretch A=[2112]\themap{A} = \begin{bmatrix} 2 & 1 \\ 1 & 2 \end{bmatrix}, the individual entries of B=P−1AP\themap{B} = P^{-1}\themap{A}P change a lot, but if one main diagonal entry goes up, the other goes down by the same amount so their sum stays at 44 (as long as u⃗1\bone{\vec{u}_1} and u⃗2\btwo{\vec{u}_2} do not line up and flatten the grid).

Changing the basis can change all four numbers in the matrix, but in every working grid the main diagonal still adds up to 4.

Key concepts

Two matrices A\themap{A} (in the standard grid) and B\themap{B} (in a new basis whose vectors form the columns of P=[b⃗1b⃗2]P = \begin{bmatrix} \bone{\vec{b}_1} & \btwo{\vec{b}_2} \end{bmatrix}) are similar when they describe the same linear map in two different coordinate systems. Whenever you convert between them, read the three-step detour from right to left based on which grid the input vector starts in:

Standard grid to a new basis (B=P−1AP\themap{B} = P^{-1}\themap{A}P)

An input vector starts in their grid, so PP translates it to our grid first, A\themap{A} transforms it, and P−1P^{-1} translates the result back. Before inverting PP, check Ab⃗1\themap{A}\bone{\vec{b}_1} and Ab⃗2\themap{A}\btwo{\vec{b}_2}. If they are already obvious combinations of b⃗1\bone{\vec{b}_1} and b⃗2\btwo{\vec{b}_2} (such as A[11]=[33]=3b⃗1+0b⃗2\themap{A}\bone{\begin{bmatrix} 1 \\ 1 \end{bmatrix}} = \bone{\begin{bmatrix} 3 \\ 3 \end{bmatrix}} = 3\bone{\vec{b}_1} + 0\btwo{\vec{b}_2}), write the columns of B\themap{B} directly without computing P−1P^{-1}. Compute P−1APP^{-1}\themap{A}P when the basis is skewed and you cannot see the combination right away.

New basis back to the standard grid (A=PBP−1\themap{A} = P\themap{B}P^{-1})

When a map is easy to write in a tilted basis (like B=[3001]\themap{B} = \begin{bmatrix} \bone{3} & \btwo{0} \\ \bone{0} & \btwo{1} \end{bmatrix}), an input vector in our standard grid is multiplied by P−1P^{-1} first to enter their grid, gets transformed by B\themap{B}, and returns to our grid through PP. Use this same detour An=PBnP−1\themap{A^n} = P\themap{B^n}P^{-1} to compute powers of A\themap{A} by raising the diagonal matrix B\themap{B} to the nnth power in the middle.

Testing similarity without solving for PP

Check properties of the map itself first: trace (sum of the main diagonal), invertibility and rank, determinant, and eigenvalues (stretch factors). Any mismatch (like trace[2112]\text{trace}\begin{bmatrix} 2 & 1 \\ 1 & 2 \end{bmatrix} =4≠trace[3112]=5{}= 4 \nomatch{\neq} \text{trace}\begin{bmatrix} 3 & 1 \\ 1 & 2 \end{bmatrix} = 5) immediately rules out similarity. If all checks match, check how the map acts on vectors rather than assuming similarity: a matrix that scales every vector equally, like I\themap{I}, satisfies P−1IP=IP^{-1}\themap{I}P = \themap{I} in every basis and can never be similar to a shear like [1101]\begin{bmatrix} 1 & 1 \\ 0 & 1 \end{bmatrix}.

6

Practice

1

In the standard grid, a linear map has the matrix A=[3122]\themap{A} = \begin{bmatrix} 3 & 1 \\ 2 & 2 \end{bmatrix}. Find the similar matrix B=P−1AP\themap{B} = P^{-1}\themap{A}P that describes this map in the new basis b⃗1=[11]\bone{\vec{b}_1} = \bone{\begin{bmatrix} 1 \\ 1 \end{bmatrix}}, b⃗2=[1−2]\btwo{\vec{b}_2} = \btwo{\begin{bmatrix} 1 \\ -2 \end{bmatrix}}.

Before finding P−1P^{-1}, multiply Ab⃗1\themap{A}\bone{\vec{b}_1} and Ab⃗2\themap{A}\btwo{\vec{b}_2} in the standard grid. How do those output vectors compare to b⃗1\bone{\vec{b}_1} and b⃗2\btwo{\vec{b}_2}?

Apply A\themap{A} to each basis vector in the standard grid and write the result in terms of b⃗1\bone{\vec{b}_1} and b⃗2\btwo{\vec{b}_2}:
Ab⃗1=[3122][11]=[44]=4b⃗1+0b⃗2\displaystyle \begin{aligned} \themap{A}\bone{\vec{b}_1} &= \begin{bmatrix} 3 & 1 \\ 2 & 2 \end{bmatrix}\bone{\begin{bmatrix} 1 \\ 1 \end{bmatrix}} = \bone{\begin{bmatrix} 4 \\ 4 \end{bmatrix}} \\ &= 4\bone{\vec{b}_1} + 0\btwo{\vec{b}_2} \end{aligned}
Ab⃗2=[3122][1−2]=[1−2]=0b⃗1+1b⃗2\displaystyle \begin{aligned} \themap{A}\btwo{\vec{b}_2} &= \begin{bmatrix} 3 & 1 \\ 2 & 2 \end{bmatrix}\btwo{\begin{bmatrix} 1 \\ -2 \end{bmatrix}} = \btwo{\begin{bmatrix} 1 \\ -2 \end{bmatrix}} \\ &= 0\bone{\vec{b}_1} + 1\btwo{\vec{b}_2} \end{aligned}
(Watch out: don't stop at the standard outputs [44]\bone{\begin{bmatrix} 4 \\ 4 \end{bmatrix}} and [1−2]\btwo{\begin{bmatrix} 1 \\ -2 \end{bmatrix}} and put them into [414−2]\begin{bmatrix} 4 & 1 \\ 4 & -2 \end{bmatrix}. The columns of B\themap{B} are the coordinates [40]\bone{\begin{bmatrix} 4 \\ 0 \end{bmatrix}} and [01]\btwo{\begin{bmatrix} 0 \\ 1 \end{bmatrix}} in the new basis b⃗1,b⃗2\bone{\vec{b}_1}, \btwo{\vec{b}_2}.)
B=[4001]\displaystyle \themap{B} = \begin{bmatrix} \bone{4} & \btwo{0} \\ \bone{0} & \btwo{1} \end{bmatrix}
2

Let A=[1213]\themap{A} = \begin{bmatrix} 1 & 2 \\ 1 & 3 \end{bmatrix} be the matrix of a map in the standard basis. Write the change-of-basis matrix PP and its inverse P−1P^{-1} for the skewed basis u⃗1=[11]\bone{\vec{u}_1} = \bone{\begin{bmatrix} 1 \\ 1 \end{bmatrix}}, u⃗2=[12]\btwo{\vec{u}_2} = \btwo{\begin{bmatrix} 1 \\ 2 \end{bmatrix}}, and compute the similar matrix B=P−1AP\themap{B} = P^{-1}\themap{A}P in this basis.

Put u⃗1\bone{\vec{u}_1} and u⃗2\btwo{\vec{u}_2} into the columns of PP, then follow the three-step detour from right to left starting with an input in the skewed basis.

Place u⃗1\bone{\vec{u}_1} and u⃗2\btwo{\vec{u}_2} into the columns of PP and invert it:
P=[1112],P−1=[2−1−11]\displaystyle P = \begin{bmatrix} \bone{1} & \btwo{1} \\ \bone{1} & \btwo{2} \end{bmatrix}, \qquad P^{-1} = \begin{bmatrix} 2 & -1 \\ -1 & 1 \end{bmatrix}
Transform the basis columns in the standard grid first:
AP=[1213][1112]=[3547]\displaystyle \themap{A}P = \begin{bmatrix} 1 & 2 \\ 1 & 3 \end{bmatrix}\begin{bmatrix} \bone{1} & \btwo{1} \\ \bone{1} & \btwo{2} \end{bmatrix} = \begin{bmatrix} \bone{3} & \btwo{5} \\ \bone{4} & \btwo{7} \end{bmatrix}
Translate the outputs back to the skewed basis by multiplying on the left by P−1P^{-1} (putting PP on the left and P−1P^{-1} on the right computes PAP−1=[−13−25]P\themap{A}P^{-1} = \begin{bmatrix} -1 & 3 \\ -2 & 5 \end{bmatrix}, which translates in the wrong direction):
B=P−1AP=[2−1−11][3547]=[2312]\displaystyle \begin{aligned} \themap{B} &= P^{-1}\themap{A}P \\ &= \begin{bmatrix} 2 & -1 \\ -1 & 1 \end{bmatrix}\begin{bmatrix} \bone{3} & \btwo{5} \\ \bone{4} & \btwo{7} \end{bmatrix} \\ &= \begin{bmatrix} \bone{2} & \btwo{3} \\ \bone{1} & \btwo{2} \end{bmatrix} \end{aligned}
P=[1112],P−1=[2−1−11],B=[2312]\displaystyle \begin{aligned} P &= \begin{bmatrix} \bone{1} & \btwo{1} \\ \bone{1} & \btwo{2} \end{bmatrix}, \\ P^{-1} &= \begin{bmatrix} 2 & -1 \\ -1 & 1 \end{bmatrix}, \\ \themap{B} &= \begin{bmatrix} \bone{2} & \btwo{3} \\ \bone{1} & \btwo{2} \end{bmatrix} \end{aligned}
3

A transformation of the plane stretches vectors along b⃗1=[11]\bone{\vec{b}_1} = \bone{\begin{bmatrix} 1 \\ 1 \end{bmatrix}} by a factor of 33 and flips vectors along the perpendicular direction b⃗2=[−11]\btwo{\vec{b}_2} = \btwo{\begin{bmatrix} -1 \\ 1 \end{bmatrix}} to their negatives. Write the matrix B\themap{B} for this map in the basis b⃗1,b⃗2\bone{\vec{b}_1}, \btwo{\vec{b}_2}, and use A=PBP−1\themap{A} = P\themap{B}P^{-1} to find its standard matrix A\themap{A}.

Since b⃗2\btwo{\vec{b}_2} is flipped to −b⃗2\btwo{-\vec{b}_2}, what is the second column of B\themap{B} in the b⃗1,b⃗2\bone{\vec{b}_1}, \btwo{\vec{b}_2} grid? Then take the detour that starts in our standard grid.

In the basis b⃗1,b⃗2\bone{\vec{b}_1}, \btwo{\vec{b}_2}, b⃗1\bone{\vec{b}_1} goes to 3b⃗1+0b⃗23\bone{\vec{b}_1} + 0\btwo{\vec{b}_2} and b⃗2\btwo{\vec{b}_2} goes to 0b⃗1−1b⃗20\bone{\vec{b}_1} - 1\btwo{\vec{b}_2} (watch out: flipping b⃗2\btwo{\vec{b}_2} means multiplying by −1-1, not 00 or +1+1):
B=[300−1]\displaystyle \themap{B} = \begin{bmatrix} \bone{3} & \btwo{0} \\ \bone{0} & \btwo{-1} \end{bmatrix}
Write PP and its inverse P−1P^{-1} (don't forget to divide by the determinant 1(1)−(−1)(1)=21(1) - (-1)(1) = 2 when finding P−1P^{-1}, or A\themap{A} will come out doubled as [2442]\begin{bmatrix} 2 & 4 \\ 4 & 2 \end{bmatrix}):
P=[1−111],P−1=[1212−1212]\displaystyle \begin{aligned} P &= \begin{bmatrix} \bone{1} & \btwo{-1} \\ \bone{1} & \btwo{1} \end{bmatrix}, \\ P^{-1} &= \begin{bmatrix} \tfrac{1}{2} & \tfrac{1}{2} \\ -\tfrac{1}{2} & \tfrac{1}{2} \end{bmatrix} \end{aligned}
Multiply A=PBP−1\themap{A} = P\themap{B}P^{-1}:
A=[1−111][300−1][1212−1212]=[313−1][1212−1212]=[1221]\displaystyle \begin{aligned} \themap{A} &= \begin{bmatrix} \bone{1} & \btwo{-1} \\ \bone{1} & \btwo{1} \end{bmatrix}\begin{bmatrix} \bone{3} & \btwo{0} \\ \bone{0} & \btwo{-1} \end{bmatrix}\begin{bmatrix} \tfrac{1}{2} & \tfrac{1}{2} \\ -\tfrac{1}{2} & \tfrac{1}{2} \end{bmatrix} \\ &= \begin{bmatrix} \bone{3} & \btwo{1} \\ \bone{3} & \btwo{-1} \end{bmatrix}\begin{bmatrix} \tfrac{1}{2} & \tfrac{1}{2} \\ -\tfrac{1}{2} & \tfrac{1}{2} \end{bmatrix} \\ &= \begin{bmatrix} 1 & 2 \\ 2 & 1 \end{bmatrix} \end{aligned}
B=[300−1],A=[1221]\displaystyle \themap{B} = \begin{bmatrix} \bone{3} & \btwo{0} \\ \bone{0} & \btwo{-1} \end{bmatrix}, \quad \themap{A} = \begin{bmatrix} 1 & 2 \\ 2 & 1 \end{bmatrix}
4

The matrix A=[3−210]\themap{A} = \begin{bmatrix} 3 & -2 \\ 1 & 0 \end{bmatrix} satisfies A=PBP−1\themap{A} = P\themap{B}P^{-1} for the basis P=[2111]P = \begin{bmatrix} \bone{2} & \btwo{1} \\ \bone{1} & \btwo{1} \end{bmatrix} (with P−1=[1−1−12]P^{-1} = \begin{bmatrix} 1 & -1 \\ -1 & 2 \end{bmatrix}) and the diagonal matrix B=[2001]\themap{B} = \begin{bmatrix} \bone{2} & \btwo{0} \\ \bone{0} & \btwo{1} \end{bmatrix}. Use the three-step detour to compute A5\themap{A^5}.

Translate to the tilted basis once, apply the diagonal stretch five times in a row, and translate back to the standard grid at the end.

Raise the diagonal matrix B\themap{B} to the 55th power in the tilted basis:
B5=[250015]=[32001]\displaystyle \themap{B^5} = \begin{bmatrix} \bone{2^5} & \btwo{0} \\ \bone{0} & \btwo{1^5} \end{bmatrix} = \begin{bmatrix} \bone{32} & \btwo{0} \\ \bone{0} & \btwo{1} \end{bmatrix}
Use A5=PB5P−1\themap{A^5} = P\themap{B^5}P^{-1} (be careful not to swap PP and P−1P^{-1} into P−1B5PP^{-1}\themap{B^5}P, which gives [6331−62−30]\begin{bmatrix} 63 & 31 \\ -62 & -30 \end{bmatrix}, and never raise the individual entries of A\themap{A} directly to the 55th power):
A5=[2111][32001][1−1−12]=[641321][1−1−12]=[63−6231−30]\displaystyle \begin{aligned} \themap{A^5} &= \begin{bmatrix} \bone{2} & \btwo{1} \\ \bone{1} & \btwo{1} \end{bmatrix}\begin{bmatrix} \bone{32} & \btwo{0} \\ \bone{0} & \btwo{1} \end{bmatrix}\begin{bmatrix} 1 & -1 \\ -1 & 2 \end{bmatrix} \\ &= \begin{bmatrix} \bone{64} & \btwo{1} \\ \bone{32} & \btwo{1} \end{bmatrix}\begin{bmatrix} 1 & -1 \\ -1 & 2 \end{bmatrix} \\ &= \begin{bmatrix} 63 & -62 \\ 31 & -30 \end{bmatrix} \end{aligned}
A5=[63−6231−30]\displaystyle \themap{A^5} = \begin{bmatrix} 63 & -62 \\ 31 & -30 \end{bmatrix}
5

Consider the three matrices M=[2412]\themap{M} = \begin{bmatrix} 2 & 4 \\ 1 & 2 \end{bmatrix}, N=[3001]\themap{N} = \begin{bmatrix} 3 & 0 \\ 0 & 1 \end{bmatrix}, and Q=[2113]\themap{Q} = \begin{bmatrix} 2 & 1 \\ 1 & 3 \end{bmatrix}. Which of these matrices, if any, are similar to one another? Justify your answer without solving for a change-of-basis matrix PP.

Compare the traces first; for the two matrices whose traces match, check whether they flatten space or are invertible.

Check the trace (sum of the main diagonal) of each matrix:
trace(M)=2+2=4,trace(N)=3+1=4,trace(Q)=2+3=5\displaystyle \begin{aligned} \text{trace}(\themap{M}) &= 2 + 2 = 4, \\ \text{trace}(\themap{N}) &= 3 + 1 = 4, \\ \text{trace}(\themap{Q}) &= 2 + 3 = 5 \end{aligned}
Because trace(Q)=5≠4\text{trace}(\themap{Q}) = 5 \nomatch{\neq} 4, Q\themap{Q} cannot be similar to M\themap{M} or N\themap{N}.
(Watch out: don't conclude that M\themap{M} and N\themap{N} are similar just because both have trace 44.) In M\themap{M}, the second column [42]\btwo{\begin{bmatrix} 4 \\ 2 \end{bmatrix}} is twice the first column [21]\bone{\begin{bmatrix} 2 \\ 1 \end{bmatrix}}, so M\themap{M} flattens the plane onto a line (rank 11, not invertible), whereas N\themap{N} is invertible (rank 22). Thus M\themap{M} and N\themap{N} cannot represent the same map either.
None of them are similar.
6

The matrices D=[3003]\themap{D} = \begin{bmatrix} 3 & 0 \\ 0 & 3 \end{bmatrix} and T=[3103]\themap{T} = \begin{bmatrix} 3 & 1 \\ 0 & 3 \end{bmatrix} both have trace 66, are both invertible with rank 22, and have the same determinant and eigenvalues. Are D\themap{D} and T\themap{T} similar? Prove your answer using the three-step detour P−1DPP^{-1}\themap{D}P.

Write D=3I\themap{D} = 3\themap{I} and simplify P−1(3I)PP^{-1}(3\themap{I})P for any invertible matrix PP.

(Watch out: passing all the quick number checks, with trace 66, rank 22, determinant 99, and eigenvalue 33, does not guarantee that D\themap{D} and T\themap{T} are similar.)
Because D=3I\themap{D} = 3\themap{I} triples every vector in space, applying the three-step detour with any basis PP gives:
P−1DP=P−1(3I)P=3P−1IP=3I=[3003]\displaystyle \begin{aligned} P^{-1}\themap{D}P &= P^{-1}(3\themap{I})P = 3P^{-1}\themap{I}P \\ &= 3\themap{I} = \begin{bmatrix} 3 & 0 \\ 0 & 3 \end{bmatrix} \end{aligned}
Since D\themap{D} has the matrix [3003]\begin{bmatrix} 3 & 0 \\ 0 & 3 \end{bmatrix} in every basis, it can never equal T=[3103]\themap{T} = \begin{bmatrix} 3 & 1 \\ 0 & 3 \end{bmatrix}.
No, they are not similar.