modelbuilder: refine existing network

Often users already have a network that they want to use as the basis of the modelbuilder. Or in this example, refine an existing network with a polygon.

import dfm_tools as dfmt
import matplotlib.pyplot as plt
plt.close("all")
import numpy as np
from meshkernel import MeshRefinementParameters, GeometryList, RefinementType

# load existing grid as xugrid.UgridDataset and take left subset
netfile = dfmt.data.fm_grevelingen_net(return_filepath=True)
uds_grid = dfmt.open_partitioned_dataset(netfile)
uds_grid = uds_grid.ugrid.sel(x=slice(None,56000))
crs = "EPSG:28992"

# depending on how the grid is written, you might have to convert mesh1d to mesh2d grid
# uds_grid = dfmt.add_network_cellinfo(uds_grid)

# convert to meshkernel object
mk_object = uds_grid.grid.meshkernel

# refine grid
pol = np.array(
    [[ 50886.71822346, 421431.47228723],
     [ 50844.41343794, 420016.93916901],
     [ 51764.54252293, 419945.01375622],
     [ 51669.35675551, 421503.39770002],
     [ 50886.71822346, 421431.47228723]]
    )
poly_geolist = GeometryList(x_coordinates=pol[:,0], y_coordinates=pol[:,1])
mk_ref_pars = MeshRefinementParameters(
    refinement_type=RefinementType.REFINEMENT_LEVELS,
    connect_hanging_nodes=True,
    smoothing_iterations=0,
    max_refinement_iterations=1,
    )
mk_object.mesh2d_refine_based_on_polygon(
    polygon=poly_geolist,
    mesh_refinement_params=mk_ref_pars,
    )

# plot the resulting grid
fig, ax = plt.subplots()
mk_object.mesh2d_get().plot_edges(ax=ax, linewidth=1)
dfmt.plot_coastlines(ax=ax, crs=crs, zorder=2, linewidth=1)
ax.plot(pol[:,0], pol[:,1], "r")

Gives:

The issue demonstrates refining an existing network through the modelbuilder workflow, using uds_grid.grid.meshkernel and mesh2d_refine_based_on_polygon with MeshRefinementParameters. Start by tracing where existing grids enter this workflow and how polygon refinement is exposed. Done means the shown network can be refined from an existing grid and produces the expected refined mesh; no source file or test is named.