Specifying flagregions for adaptive refinement

New in Version 5.7.0.

AMRClaw and GeoClaw version 5.6.1 (and earlier) allow specifying rectangular refinement regions (see Specifying AMR regions) in setrun.py, in the form of a list that is appended to rundata.regiondata.regions:

rundata.regiondata.regions.append([minlevel,maxlevel,t1,t2,x1,x2,y1,y2])

This is a region that is active from time t1 to t2 over the spatial extent [x1,x2,y1,y2].

Starting in v5.7.0 of AMRClaw/GeoClaw we now support a new approach to specifying regions that are now called flagregions for more clarity regarding what they are used for. The new data structure also supports simple rectangles and so should ultimately replace regions in both AMRClaw and GeoClaw, but currently you can mix and match.

The new way of specifying a flag region in setrun.py is to first define an object flagregion of class clawpack.amrclaw.data.FlagRegion, set various attributes of this object (including minlevel, maxlevel, t1, t2, and a spatial extent), and then append this object to the list rundata.flagregiondata.flagregions.

Here is how you would specify a simple rectangle as above in the new style, chosen to cover the entire spatial domain and to allow only 1 level everywhere (which might be supplemented by other regions where more levels are allowed):

x1,x2,y1,y2 = [rundatat.clawdata.lower[0], rundatat.clawdata.upper[0],
               rundatat.clawdata.lower[1], rundatat.clawdata.upper[1]]

from clawpack.amrclaw.data import FlagRegion
flagregion = FlagRegion(num_dim=2)  # so far only 2D supported
flagregion.name = 'Region_domain'
flagregion.minlevel = 1
flagregion.maxlevel = 1
flagregion.t1 = 0.
flagregion.t2 = 1e9
flagregion.spatial_region_type = 1  # Rectangle
flagregion.spatial_region = [x1,x2,y1,y2]
rundata.flagregiondata.flagregions.append(flagregion)

Note that flagregion.spatial_region_type == 1 indicates that the flagregion is a rectangle.

Using ruled rectangles as flagregions

In addition to simple rectangles, more general ruled rectangles can also be used as flagregions. These are a restricted set of polygons for which it is easy to test if a point is inside or outside, as described in more detail in Ruled Rectangles.

To specify a ruled rectangle, use flagregion.spatial_region_type == 2 and provide a path to a data file that describes the ruled rectangle. For simple ruled rectangles the code to create the data file can also be included in setrun.py.

Here is an example where a simple ruled rectangle is defined and used as a flagregion. In this case the flagregion is a trapezoid with vertices \((1,3),~ (1,6),~ (2,4),~ (2,7)\):

from clawpack.amrclaw.data import FlagRegion
flagregion = FlagRegion(num_dim=2)
flagregion.name = 'Region_Trapezoid'
flagregion.minlevel = 2
flagregion.maxlevel = 3
flagregion.t1 = 0.
flagregion.t2 = 1e9
flagregion.spatial_region_type = 2  # Ruled Rectangle
flagregion.spatial_region_file = \
    os.path.abspath('RuledRectangle_Trapezoid.data')
rundata.flagregiondata.flagregions.append(flagregion)

# code to make RuledRectangle_Trapezoid.data:
from clawpack.amrclaw import region_tools
rr = region_tools.RuledRectangle()
rr.method = 1 # piecewiselinear edges between s values
rr.ixy = 'x'  # so s refers to x, lower & upper are limits in y
rr.s = np.array([1,2])
rr.lower = np.array([3,6])
rr.upper = np.array([4,7])
rr.write('RuledRectangle_Trapezoid.data')  # creates data file

See the setrun.py file in $CLAW/amrclaw/examples/advection_2d_flagregions for additional examples.

See Creating an AMR flag region for a more complex example where a ruled rectangle is defined that covers a set of fgmax points (see Fixed grid monitoring) defined with the Marching Front algorithm.