Create a plan | Create an approval record | Execute with a ledger | Sample completed patches

Normalize incoming metadata | Retitle changed records | Compare duplicate candidates | Map Crossref metadata | Build a cross-library move

Query SQLite read-only | Query the optional semantic sidecar | Exercise the attachment protocol offline | Keep evidence in the correct plane

Configure API access | Read API resources | Use local and semantic data | Build a plan

Choose the right API layer | Notation and cells | Fitting and OOF scores | Unseen-row ensembles are cell-local | Independent OOF residual reconstruction | Signed residual offsets | Two distinct selection rules | Optimism diagnostic | Executable residual-offset example | References

Responses, row classes, and companion values | Jacobi recurrence | Asymmetric Gauss--Seidel recurrence | Companion-importance normalization | Fixed and shadow gates | Active-set history | Change score and stopping policy | Per-response iteration stitch | Interpretation and limits | References

Symbols and input pools | Optional augmented-data warm start | Range-ratio promotion score | Key budget and pseudo-label rows | OOF acceptance gauge | State transitions and return value | Final re-emission boundary | Limitations | References

Three questions, three helpers | Minimum data contract | Diagnose and normalize numeric text | Create the three model-input products | When to use each helper | Important limits and side effects

Direct helper or runner | Runner contract | Minimal batch pattern | Plotting boundary | Output convention

Expected file layout | Provider formats supported | Extraction pipeline | Indexing tables | Project-owned record catalog | Composing with the processing core | Audit helpers | Maintenance | Notes

Input and output | Input | Output | SDOF model | Discretisation via matrix exponential | How PSA / PSV / SD are built | D50 and D100 horizontal spectra | Notes (limitations and corner cases)

Synthetic input | Run AT2TS | Intensity measures | Response spectrum | IMF decomposition | Where to go next

Method overview | Prerequisites | The site JSON | Step 1: spectral preselection | Seismological screening | Step 2: match | Step 3: acceptance diagnostics | The in-memory API | Composition with the other runners

Why this matters | Fundamental period of an inhomogeneous truncated shear beam | Ishihara stiffness model | Synthetic profile generation | Site-class utilities | Site amplification models | Application | References

Scope | The four workflows | Workflow 1 — Dynamic site response | Workflow 2 — Hazard import | Workflow 3 — Displacement curves | Workflow 4 — Seismic coefficient | See also

Overview | Three-stage structure | Stage 1 — Site amplification (fitSaF, fitSaFMixture) | Stage 2 — Displacement curve (fitDnCurve) | Stage 3 — Coefficient inversion (invertDnDraws) | Displacement models | Key assumptions | Operational defaults | References

What is an IMF? | Input and grouping | Available engines | VMD (method = "vmd", default) | EMD (method = "emd") | EEMD (method = "eemd") | Output | Per-IMF metrics | Reconstruction filters | Composing with AT2TS / VT2TS / DT2TS | References

Expected input | Units | Internal unit conventions | Implemented measures | Long and wide output | Acceleration (ID = "AT") | Velocity (ID = "VT") | Displacement (ID = "DT") | What it does not compute | References

Data formats | Input (wide) | Outputs | Time regularisation | Internal STFT Strategy Selection | STFT audit: auditSTFT() | STFT / OLA filtering: .ffilter() | Integration: .integrate() | Differentiation: .derivate() | Anti-alias resampling: .resample() | Edge tapering: .taperA() | Notes (parameter audit) | References

Overview | Example dataset | Parameters | Workflow 3 — Displacement curve Dn(ky) | Workflow 4 — Seismic coefficient kmax(d*) | Next steps