Measuring "in the X basis" on hardware that only measures Z takes one Hadamard. Behind that trick is a general principle: a change of orthonormal basis is a unitary, and descriptions related by a unitary are the same physics. So when are two operators really the same up to coordinates?
What you’ll learn
Derive how coordinates and matrices transform under a change of orthonormal basis, and distinguish passive from active transformations.
Decide unitary equivalence: by spectra for normal operators, by trace invariants (Specht) in general.
Implement measurements in rotated bases by conjugating with unitaries.
Move between the Schrödinger, Heisenberg and interaction pictures.
The transition matrix
Let {∣ei⟩} and {∣fj⟩} be orthonormal bases of Cd. The transition matrixWij=⟨ei∣fj⟩ has the new basis vectors, in old coordinates, as its columns; it is unitary because both bases are orthonormal. Inserting resolutions of the identity does everything:
DerivationCoordinates and matrices under a change of basis
Passive versus active. Rewriting the same state in new coordinates (passive) uses W†; physically rotating the state in fixed coordinates, ∣ψ⟩↦W∣ψ⟩ (active), uses W. Predictions are unchanged when states and observables transform together:
For a qubit, Clifford gates relate the standard bases. With W=H (columns ∣±⟩): HZH=X and HXH=Z. With W=SH (columns ∣±i⟩): (SH)†Y(SH)=Z. To measure an observable A=WZW† on a Z-only device, apply W† and then measure Z; outcome 0 then certifies W∣0⟩.
Unitary equivalence: what coordinates cannot change
Similarity, B=S−1AS with S merely invertible, is weaker. J1=(1011) and J2=(1021) are similar (via S=diag(2,1)), but TrJ1†J1=3=6=TrJ2†J2, so no change of orthonormal basis relates them. For normal matrices the two notions coincide.
DerivationCommuting normal operators share an eigenbasis
Pk=l=k∏ak−alA−alI⟹BPk=PkB
Let A,B be normal with AB=BA and A=∑kakPk. Each spectral projector is a polynomial in A (Lagrange interpolation), so B commutes with it:
B(ranPk)⊆ranPk
So B (and, taking adjoints, B†) maps each eigenspace of A into itself, and its restriction there is still normal:
A=n∑an∣n⟩⟨n∣,B=n∑bn∣n⟩⟨n∣
Diagonalize B by an orthonormal basis inside each eigenspace. Together these diagonalize both, which is why commuting observables can be measured jointly:
Check your understanding
Which pair is unitarily equivalent?
X and S
Z and 2Z
X and Z
(1011) and I
Show the answer
C.X and Z
X and Z are normal with the same spectrum {+1,−1}; explicitly HZH=X.
Pictures: time as a change of basis
Time evolution is a time-dependent unitary, so it can act on states or on observables. In the Schrödinger picture∣ψ(t)⟩=U(t)∣ψ⟩ and observables stay fixed. In the Heisenberg picture states stay fixed and AH(t)=U(t)†AU(t). Predictions agree: ⟨ψ(t)∣A∣ψ(t)⟩=⟨ψ∣AH(t)∣ψ⟩.
DerivationThe Heisenberg equation and Larmor precession
dtdAH(t)=eiHt(iHA−iAH)e−iHt=i[H,AH(t)]
For time-independent H, U(t)=e−iHt (ℏ=1). Differentiate AH(t)=eiHtAe−iHt:
X˙H=i2ω[Z,XH]=−ωYH,Y˙H=ωXH
Take H=2ωZ (so ZH=Z) and use [Z,X]=2iY, [Z,Y]=−2iX:
XH(t)=Xcosωt−Ysinωt,YH(t)=Ycosωt+Xsinωt
Solve. The observables rotate while the state stays put; in the Schrödinger picture the Bloch vector precesses instead. The two descriptions differ by the change of basis U(t).
The interaction picture splits H=H0+V and moves only the easy part into the basis: ∣ψI(t)⟩=eiH0t∣ψ(t)⟩ evolves under VI(t)=eiH0tVe−iH0t. For a driven qubit this is the rotating frame, where the rotating-wave approximation becomes visible (see Rabi Oscillations & the Rotating Frame).
Think first If states evolve as ρ↦UρU†, how must observables transform so that Tr(ρA) is unchanged when you switch pictures?
By cyclicity, Tr(UρU†A)=Tr(ρU†AU), so the Heisenberg-picture observable is U†AU. For a general channelΦ this becomes the dual map Φ†, defined by Tr(Φ(ρ)A)=Tr(ρΦ†(A)).
Canonical forms and complementary bases
Unitary equivalence also yields canonical forms. Schur: every A=UTU† with T upper triangular. Singular value decomposition: every A=UΣV†, with one change of basis on the input and another on the output; this is behind the Schmidt decomposition. Clifford conjugation: the Clifford group consists exactly of the unitaries that map Pauli operators to Pauli operators up to phase, i.e. that permute the Pauli coordinate axes.
Two bases are mutually unbiased if ∣⟨ei∣fj⟩∣2=d1 for all i,j: a basis vector of one gives uniformly random outcomes in the other. The Z, X and Y bases of a qubit are pairwise unbiased, and the Fourier matrix Fjk=ωjk/d maps the computational basis to an unbiased one in every d. At most d+1 bases of Cd can be pairwise unbiased, and d+1 are achieved when d is a prime power (Wootters–Fields).
Key takeaways
A change of orthonormal basis is the unitary Wij=⟨ei∣fj⟩: coordinates become W†ψ and matrices W†AW.
To measure A=WZW† with a Z detector, apply W† first; for Y that is S†, then H.
Normal operators are unitarily equivalent iff their spectra agree; in general Specht's trace-word invariants decide, and similarity is weaker.
Commuting normal operators share an orthonormal eigenbasis, so commuting observables are jointly measurable.
The Schrödinger, Heisenberg and interaction pictures are the same physics in time-dependent coordinates, with A˙H=i[H,AH].
Check yourself
In the interactive lesson these are scored with your confidence, and what you miss comes back for review.
Question 1
With Wij=⟨ei∣fj⟩, what is the matrix of A in the new basis {∣fj⟩}?
WA(e)W†
W†A(e)W
WTA(e)W
W−1A(e)W−1
Show the answer
B.W†A(e)W
⟨fj∣A∣fk⟩=∑ilWij∗AilWlk=(W†AW)jk.
Question 2
Two n×n Hermitian matrices are unitarily equivalent if and only if they have the same...
trace and determinant
rank
eigenvalues, counted with multiplicity
diagonal entries in the computational basis
Show the answer
C. eigenvalues, counted with multiplicity
Hermitian matrices are normal. If both have spectrum D (with multiplicity), then A=U†DU and B=V†DV, so B=(U†V)†A(U†V).
Question 3
You can only measure in the Z basis. Which gates, in time order, should precede the measurement so that outcome 0 certifies ∣+i⟩ and outcome 1 certifies ∣−i⟩?
S†, then H
H, then S†
H only
S† only
Show the answer
A.S†, then H
HS†∣+i⟩=H∣+⟩=∣0⟩ and HS†∣−i⟩=H∣−⟩=∣1⟩. This is W† for W=SH, whose columns are ∣±i⟩: undo the basis change, then measure Z.
Question 4
Are J1=(1011) and J2=(1021) unitarily equivalent?
Yes: they have the same eigenvalues
Yes: they are similar, which is the same thing
No: their traces differ
No: TrJ1†J1=3 but TrJ2†J2=6
Show the answer
D. No: TrJ1†J1=3 but TrJ2†J2=6
TrA†A=∥A∥22 is a unitary invariant, and it differs. They are similar via S=diag(2,1), but similarity allows non-orthonormal bases.
Further reading
R. A. Horn & C. R. Johnson, Matrix Analysis, ch. 2 (Cambridge, 2nd ed., 2013).
J. J. Sakurai & J. Napolitano, Modern Quantum Mechanics, ch. 2 (Cambridge, 3rd ed., 2020).
W. K. Wootters & B. D. Fields, "Optimal state-determination by mutually unbiased measurements," Ann. Phys. 191, 363 (1989).
T. Durt, B.-G. Englert, I. Bengtsson & K. Życzkowski, "On mutually unbiased bases," Int. J. Quantum Inf. 8, 535 (2010).
D. Janzing, P. Wocjan & T. Beth, "'Non-identity check' is QMA-complete," Int. J. Quantum Inf. (2005).