Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Add function for converting MPAS meshes to viz triangles #319

Merged
merged 1 commit into from Jul 28, 2020
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Jump to
Jump to file
Failed to load files.
Diff view
Diff view
1 change: 1 addition & 0 deletions conda_package/mpas_tools/viz/__init__.py
@@ -0,0 +1 @@
from mpas_tools.viz.mesh_to_triangles import mesh_to_triangles
147 changes: 147 additions & 0 deletions conda_package/mpas_tools/viz/mesh_to_triangles.py
@@ -0,0 +1,147 @@
import xarray
import numpy


def mesh_to_triangles(dsMesh, periodicCopy=False):
"""
Construct a dataset in which each MPAS cell is divided into the triangles
connecting pairs of adjacent vertices to cell centers.

Parameters
----------
dsMesh : xarray.Dataset
An MPAS mesh

periodicCopy : bool, optional
Whether to make a periodic copy of triangles that cross -180/180 degrees
longitude. This is helpful when plotting triangles in a lon/lat space.

Returns
-------
dsTris : xarray.Dataset
A dataset that defines triangles connecting pairs of adjacent vertices
to cell centers as well as the cell index that each triangle is in and
cell indices and weights for interpolating data defined at cell centers
to triangle nodes. ``dsTris`` includes variables ``triCellIndices``,
the cell that each triangle is part of; ``nodeCellIndices`` and
``nodeCellWeights``, the indices and weights used to interpolate from
MPAS cell centers to triangle nodes; Cartesian coordinates ``xNode``,
``yNode``, and ``zNode``; and ``lonNode``` and ``latNode`` in radians.
``lonNode`` is guaranteed to be within 180 degrees of the cell center
corresponding to ``triCellIndices``. Nodes always have a
counterclockwise winding.

"""
nVerticesOnCell = dsMesh.nEdgesOnCell.values
verticesOnCell = dsMesh.verticesOnCell.values - 1
cellsOnVertex = dsMesh.cellsOnVertex.values - 1

kiteAreasOnVertex = dsMesh.kiteAreasOnVertex.values

nTriangles = numpy.sum(nVerticesOnCell)

maxEdges = dsMesh.sizes['maxEdges']
nCells = dsMesh.sizes['nCells']
if dsMesh.sizes['vertexDegree'] != 3:
raise ValueError('mesh_to_triangles only supports meshes with '
'vertexDegree = 3')

# find the third vertex for each triangle
nextVertex = -1*numpy.ones(verticesOnCell.shape, int)
for iVertex in range(maxEdges):
valid = iVertex < nVerticesOnCell
invalid = numpy.logical_not(valid)
verticesOnCell[invalid, iVertex] = -1
nv = nVerticesOnCell[valid]
cellIndices = numpy.arange(0, nCells)[valid]
iNext = numpy.where(iVertex < nv - 1, iVertex + 1, 0)
nextVertex[:, iVertex][valid] = verticesOnCell[cellIndices, iNext]

valid = verticesOnCell >= 0
verticesOnCell = verticesOnCell[valid]
nextVertex = nextVertex[valid]

# find the cell index for each triangle
triCellIndices, _ = numpy.meshgrid(numpy.arange(0, nCells),
numpy.arange(0, maxEdges),
indexing='ij')
triCellIndices = triCellIndices[valid]

# find list of cells and weights for each triangle node
nodeCellIndices = -1*numpy.ones((nTriangles, 3, 3), int)
nodeCellWeights = numpy.zeros((nTriangles, 3, 3))

# the first node is at the cell center, so the value is just the one from
# that cell
nodeCellIndices[:, 0, 0] = triCellIndices
nodeCellWeights[:, 0, 0] = 1.

# the other 2 nodes are associated with vertices
nodeCellIndices[:, 1, :] = cellsOnVertex[verticesOnCell, :]
nodeCellWeights[:, 1, :] = kiteAreasOnVertex[verticesOnCell, :]
nodeCellIndices[:, 2, :] = cellsOnVertex[nextVertex, :]
nodeCellWeights[:, 2, :] = kiteAreasOnVertex[nextVertex, :]

weightSum = numpy.sum(nodeCellWeights, axis=2)
for iNode in range(3):
nodeCellWeights[:, :, iNode] = nodeCellWeights[:, :, iNode]/weightSum

dsTris = xarray.Dataset()
dsTris['triCellIndices'] = ('nTriangles', triCellIndices)
dsTris['nodeCellIndices'] = (('nTriangles', 'nNodes', 'nInterp'),
nodeCellIndices)
dsTris['nodeCellWeights'] = (('nTriangles', 'nNodes', 'nInterp'),
nodeCellWeights)

# get Cartesian and lon/lat coordinates of each node
for prefix in ['x', 'y', 'z', 'lat', 'lon']:
outVar = '{}Node'.format(prefix)
cellVar = '{}Cell'.format(prefix)
vertexVar = '{}Vertex'.format(prefix)
coord = numpy.zeros((nTriangles, 3))
coord[:, 0] = dsMesh[cellVar].values[triCellIndices]
coord[:, 1] = dsMesh[vertexVar].values[verticesOnCell]
coord[:, 2] = dsMesh[vertexVar].values[nextVertex]
dsTris[outVar] = (('nTriangles', 'nNodes'), coord)

# nothing obvious we can do about triangles containing the poles

# make sure node longitudes are within 180 degrees of the cell center
lonNode = dsTris.lonNode.values
lonCell = lonNode[:, 0]
copyEast = numpy.zeros(lonCell.shape, bool)
copyWest = numpy.zeros(lonCell.shape, bool)
for iNode in [1, 2]:
mask = lonNode[:, iNode] - lonCell > numpy.pi
copyEast = numpy.logical_or(copyEast, mask)
lonNode[:, iNode][mask] = lonNode[:, iNode][mask] - 2*numpy.pi
mask = lonNode[:, iNode] - lonCell < -numpy.pi
copyWest = numpy.logical_or(copyWest, mask)
lonNode[:, iNode][mask] = lonNode[:, iNode][mask] + 2*numpy.pi
if periodicCopy:
dsNew = xarray.Dataset()
eastIndices = numpy.nonzero(copyEast)[0]
westIndices = numpy.nonzero(copyWest)[0]
triIndices = numpy.append(numpy.append(numpy.arange(0, nTriangles),
eastIndices), westIndices)

dsNew['triCellIndices'] = ('nTriangles', triCellIndices[triIndices])
dsNew['nodeCellIndices'] = (('nTriangles', 'nNodes', 'nInterp'),
nodeCellIndices[triIndices, :, :])
dsNew['nodeCellWeights'] = (('nTriangles', 'nNodes', 'nInterp'),
nodeCellWeights[triIndices, :, :])
for prefix in ['x', 'y', 'z', 'lat']:
outVar = '{}Node'.format(prefix)
coord = dsTris[outVar].values
dsNew[outVar] = (('nTriangles', 'nNodes'), coord[triIndices, :])

coord = dsTris['lonNode'].values[triIndices, :]
eastSlice = slice(nTriangles, nTriangles+len(eastIndices))
coord[eastSlice, :] = coord[eastSlice, :] + 2*numpy.pi
westSlice = slice(nTriangles+len(eastIndices),
nTriangles + len(eastIndices) + len(westIndices))
coord[westSlice, :] = coord[westSlice, :] - 2 * numpy.pi
dsNew['lonNode'] = (('nTriangles', 'nNodes'), coord)
dsTris = dsNew

return dsTris