zoomy_core.systemmodel.column#
The model’s vertical column maps in NumPy: interpolate_to_3d (a cell state -> the canonical
[b, h, u, v, w, p] column at nz heights) and its inverse project_from_3d (a sampled
column -> the state), the exchange of a coupling to a resolved free-surface solver (VOF).
The heights are the cell-centre nodes zeta_k = (k + 1/2) / nz the model bakes into
project_from_3d (SME(project_nz=nz)), so project(interpolate(q)) is the model’s own
quadrature of its profile and a peer sampling its column on the same nodes needs no
interpolation.
Module Contents#
Classes#
|
Data#
API#
- zoomy_core.systemmodel.column.FIELDS#
(‘b’, ‘h’, ‘u’, ‘v’, ‘w’, ‘p’)
- class zoomy_core.systemmodel.column.ColumnMap(sm, nz, aux=None)#
interpolate(Q, Qaux, p)andproject(C)of a SystemModel (or NSM)smonnznodes.Qis(n_state, n),Qaux(n_aux, n)in the aux order ofaux(defaultsm’s; a NumericalSystemModel’s appends its own rows, which the maps do not read), a columnC(6, nz, n)inFIELDSorder.Initialization
- interpolate(Q, Qaux, p)#
The column
(6, nz, n)of the statesQwith auxQaux, parametersp.
- project(C)#
The states
(n_state, n)of the columnsC(6, nz, n)(depth-constant fields from the first node).