This checkpoint pulls the whole module together. You will take a Hermitian observable, find its
spectrum and eigenvectors, and rebuild it from its spectral decomposition — the round trip that the
spectral theorem promises always works. If your eigenvalues and eigenvectors are right, the rebuilt
operator reproduces the original exactly.
The task
Diagonalize the observable
A=(1221).
It is real and symmetric, hence Hermitian, so the spectral theorem guarantees real eigenvalues and an
orthonormal eigenbasis.
Step 1 — eigenvalues
Use the trace–determinant form of the characteristic equation. Here trA=2 and
detA=(1)(1)−(2)(2)=−3, so
λ2−2λ−3=(λ−3)(λ+1)=0⟹λ1=3,λ2=−1.
Both real, as required. Their sum is 2=trA and product is −3=detA — a quick
consistency check.
Step 2 — eigenvectors
Solve (A−λ1)∣v⟩=0 for each eigenvalue.
For λ1=3: (−222−2)∣v⟩=0 gives v1=v2,
so ∣v1⟩=21(11).
For λ2=−1: (2222)∣v⟩=0 gives v1=−v2,
so ∣v2⟩=21(1−1).
The two eigenvectors are orthogonal, ⟨v1∣v2⟩=21(1−1)=0, exactly as the
spectral theorem demands for distinct eigenvalues of a Hermitian operator. Each is normalized to unit
length.
Step 3 — reconstruct from the spectral decomposition
The spectral theorem says
A=λ1∣v1⟩⟨v1∣+λ2∣v2⟩⟨v2∣.
The two outer products are
∣v1⟩⟨v1∣=21(1111),∣v2⟩⟨v2∣=21(1−1−11).
Weighting by the eigenvalues and adding,
3⋅21(1111)+(−1)⋅21(1−1−11)=(1221)=A.
The reconstruction matches A entry by entry. In particular the (0,0) entry is 3⋅21+(−1)⋅21=1, which equals A00 — the value you will return as a check.
Try it
Carry out the three steps above in code: find the eigenvalues, build normalized eigenvectors, form the
spectral decomposition, and return the reconstructed (0,0) entry of A. A correct diagonalization
returns 1.
Run your code to see the quantum state.
Sign in on the full site to ask questions and join the discussion.