---

name: "core-objects" description: "QuTiP quantum-object construction, algebra, measurement, and metric workflows." metadata: disco-role: operating disable-model-invocation: true license: BSD 3-Clause

Core objects

Use this subskill when the task is about QuTiP's fundamental quantum-object layer: Qobj, state and operator construction, tensor products, superoperators, random states, measurements, metrics, entropy, partial transpose, and gate helpers.

Read this subskill when the prompt mentions

• Qobj, basis, ket2dm, qeye, destroy, create, sigmax, sigmaz, tensor
• density matrices, kets, bras, operator dimensions, or tensor structure
• measure_observable, measurement_statistics_observable, or POVM-style measurements
• fidelity, tracedist, bures_dist, average_gate_fidelity, or similar metrics
• random states/operators, superoperators, partial trace, partial transpose, or entropy

What to do first

1. Decide the Hilbert-space dimensions before building objects.
2. Create the simplest valid Qobj or constructor result that matches the physics.
3. Check dims, shape, isherm, isket, isbra, isoper, or superrep before doing anything more complicated.
4. If the task includes evolution, coefficients, or plotting, route to a different subskill instead of expanding this one.

Core workflow

• Use the state and operator constructors from qutip.core.states and qutip.core.operators.
• Compose systems with tensor and super_tensor when tensor structure matters.
• Convert between kets and density matrices with ket2dm.
• Use measurement helpers only when the operator/state dimensions are already compatible.
• Use metrics and entropy helpers after the state or channel is built and sanity-checked.

Typical success signals

• The object prints the expected tensor dimensions.
• The object type matches the intended physics role.
• Measurement probabilities sum to one.
• Metrics like fidelity and trace distance return finite values in the expected range.

Boundaries

Use this subskill for object validity and algebra. Do not use it as the primary route for:

• Time evolution, collapse operators, solver options, or time-dependent Hamiltonians; use dynamics-and-solvers.
• Bath models, PIQS, HEOM, or transfer tensors; use specialized-open-systems.
• Figure creation, tomography plots, or object serialization; use analysis-and-io.

Answer shape

When responding from this subskill, give:

1. The concrete constructor or helper API to use.
2. A minimal object-building code snippet.
3. The dimension/type checks that prove the object is valid.
4. The next route if the user wants to evolve, plot, or save the object.

Validation hints

• For tensor workflows, print dims before and after tensor or ptrace.
• For measurements, check that all projectors/observables share the state dimensions.
• For metrics, check normalization and object type before trusting the number.

Reference files

• references/api-reference.md for constructors, object properties, and the most-used helper families.
• references/workflows.md for small composition, measurement, and comparison recipes.
• references/troubleshooting.md for dims mismatch, type mismatch, and measurement errors.

Helper script

• scripts/core_smoke.py runs a tiny object-construction, tensor, measurement, and metric smoke check.