source: production/sydney_2006/run_sydney_smf_polyingo.py @ 3347

"""Script for running a tsunami inundation scenario for Sydney, NSW, Australia.

Source data such as elevation and boundary data is assumed to be available in
directories specified by project.py
The output sww file is stored in project.outputdir 

The scenario is defined by a triangular mesh created from project.polygon,
the elevation data and a simulated submarine landslide.

Ole Nielsen and Duncan Gray, GA - 2005 and Adrian Hitchman and Jane Sexton, GA - 2006
"""


#-------------------------------------------------------------------------------# Import necessary modules
#-------------------------------------------------------------------------------

# Standard modules
import os
import time

# Related major packages
from pyvolution.shallow_water import Domain, Reflective_boundary, Time_boundary, Dirichlet_boundary
from pyvolution.data_manager import convert_dem_from_ascii2netcdf, dem2pts
from pyvolution.combine_pts import combine_rectangular_points_files 
from pyvolution.pmesh2domain import pmesh_to_domain_instance
from pyvolution.quantity import Quantity
from geospatial_data import *
from utilities.polygon import inside_polygon
from Numeric import allclose

# Application specific imports
import project                 # Definition of file names and polygons
from smf import slump_tsunami  # Function for submarine mudslide 


#-------------------------------------------------------------------------------
# Preparation of topographic data
#
# Convert ASC 2 DEM 2 PTS using source data and store result in source data
# Do for coarse and fine data
# Fine pts file to be clipped to area of interest
#-------------------------------------------------------------------------------
# filenames
coarsedemname = project.coarsedemname
finedemname = project.finedemname


# coarse data
convert_dem_from_ascii2netcdf(coarsedemname, use_cache=True, verbose=True)
#dem2pts(coarsedemname, use_cache=True, verbose=True)
dem2pts(coarsedemname,
        easting_min=project.eastingmin,
        easting_max=project.eastingmax,
        northing_min=project.northingmin,
        northing_max= project.northingmax,
        use_cache=True, 
        verbose=True)

# fine data (clipping the points file to smaller area)
convert_dem_from_ascii2netcdf(finedemname, use_cache=True, verbose=True)
dem2pts(finedemname,
        easting_min=project.eastingmin,
        easting_max=project.eastingmax,
        northing_min=project.northingmin,
        northing_max= project.northingmax,
        use_cache=True, 
        verbose=True)


# combining the coarse and fine data
combine_rectangular_points_files(project.finedemname + '.pts',
                                 project.coarsedemname + '.pts',
                                 project.combineddemname + '.pts')

#-------------------------------------------------------------------------------                                 
# Create the triangular mesh based on overall clipping polygon with a tagged
# boundary and interior regions defined in project.py along with
# resolutions (maximal area of per triangle) for each polygon
#-------------------------------------------------------------------------------
#from pmesh.create_mesh import create_mesh_from_regions
from pmesh.mesh_interface import create_mesh_from_regions

meshname = project.meshname4+'.msh'

interior_res1 = 1000
interior_res2 = 1315

# Read in files containing polygon points (generated by GIS)
# and construct list of interior polygons into interior_regions.

def get_polygon_from_file(filename):
    """ Function to read in output from GIS determined polygon
    """
    fid = open(filename)
    lines = fid.readlines()
    fid.close()
    
    polygon = []
    for line in lines[1:]:
       fields = line.split(',')
       x = float(fields[1])
       y = float(fields[2])
       polygon.append([x, y])
        
    return polygon

num_polygons = 4
fileext = '.csv'
filename = project.polygonptsfile

interior_regions = []
bounding_polygon = project.diffpolygonall
count = 0
for p in range(3, num_polygons+1):
    thefilename = filename + str(p) + fileext
    print 'reading in polygon points', thefilename
    interior_polygon = get_polygon_from_file(thefilename)
    interior_regions.append([interior_polygon, interior_res1])
    n = len(interior_polygon)
    # check interior polygon falls in bounding polygon
    if len(inside_polygon(interior_polygon, bounding_polygon,
                   closed = True, verbose = False)) <> len(interior_polygon):
        print 'WARNING: interior polygon %d is outside bounding polygon' %(p)
        count += 1
    # check for duplicate points in interior polygon
    
print 'number of interior polygons: ', len(interior_regions)
if count == 0: print 'interior polygons OK'

# original
#interior_regions = []
#interior_regions = [[project.harbour_polygon_2, interior_res1],
#                    [project.botanybay_polygon_2, interior_res1]]

# finer set
#interior_regions = [[project.newpoly1, interior_res1],
#                    #[project.parrariver, interior_res1], #remove for second run
#                    [project.south1, interior_res1],
#                    [project.finepolymanly, interior_res2],
#                    [project.finepolyquay, interior_res2]]


#FIXME: Fix caching of this one once the mesh_interface is ready
from caching import cache


#_ = cache(create_mesh_from_regions,
#          project.diffpolygonall,
#          {'boundary_tags': {'bottom': [0], 
#                             'right1': [1], 'right0': [2],
#                             'right2': [3], 'top': [4], 'left1': [5],
#                             'left2': [6], 'left3': [7]},
#           'maximum_triangle_area': 150000,
#           'filename': meshname,           
#           'interior_regions': interior_regions},
#          verbose = True)

# for demo construction
_ = cache(create_mesh_from_regions,
          project.demopoly,
          {'boundary_tags': {'bottom': [0], 'right': [1],
                             'top': [2], 'left': [3]},
           'maximum_triangle_area': 100000,
           'filename': meshname,           
           'interior_regions': interior_regions},
          verbose = True)

#mesh_geo_reference:
#-------------------------------------------------------------------------------                                 
# Setup computational domain
#-------------------------------------------------------------------------------                                 

domain = pmesh_to_domain_instance(meshname, Domain,
                                  use_cache = True,
                                  verbose = True)

print 'Number of triangles = ', len(domain)
print 'The extent is ', domain.get_extent()
print domain.statistics()

domain.set_name(project.basename4)
domain.set_datadir(project.outputdir)
domain.set_quantities_to_be_stored(['stage', 'xmomentum', 'ymomentum'])


#-------------------------------------------------------------------------------
# Set up scenario (tsunami_source is a callable object used with set_quantity)
#-------------------------------------------------------------------------------

tsunami_source = slump_tsunami(length=30000.0,
                               depth=400.0,
                               slope=6.0,
                               thickness=176.0, 
                               radius=3330,
                               dphi=0.23,
                               x0=project.slump_origin[0], 
                               y0=project.slump_origin[1], 
                               alpha=0.0, 
                               domain=domain)

#-------------------------------------------------------------------------------                                 
# Setup initial conditions
#-------------------------------------------------------------------------------

G = Geospatial_data(project.test_pts, project.test_elev) #points, attributes
domain.set_quantity('stage', 0) #tsunami_source)
domain.set_quantity('friction', 0.0) # supplied by Benfield, initial value 0.03
domain.set_quantity('elevation', 
                    filename = project.coarsedemname + '.pts',
                    use_cache = True,
                    verbose = True)

from pyvolution.least_squares import fit_to_mesh_file, DEFAULT_ALPHA 

#Add elevation data to the mesh
#fit_to_mesh_file(mesh_file, point_file, mesh_output_file, DEFAULT_ALPHA,
#                 verbose=True, expand_search=True, precrop=True)
#pointname = project.coarsedemname + '.pts'
#mesh_elevname = project.meshelevname
#cache(fit_to_mesh_file,(meshname,
#                 pointname,
#                 mesh_elevname,
#                 DEFAULT_ALPHA),
#      {'verbose': True,
#       'expand_search': True,
#       'precrop': True}
#      ,dependencies = [meshname, pointname]
#      #,evaluate = True      
#      ,verbose = False
#      )

#-------------------------------------------------------------------------------                                 
# Setup boundary conditions (all reflective)
#-------------------------------------------------------------------------------

print 'Available boundary tags', domain.get_boundary_tags()

Br = Reflective_boundary(domain)
Bd = Dirichlet_boundary([0, 0, 0])

# for demo
domain.set_boundary( {'bottom': Bd, 'right': Bd, 'left': Bd, 'top': Bd} )
#domain.set_boundary( {'bottom': Br, 'right1': Br, 'right0': Br,
#                      'right2': Br, 'top': Br, 'left1': Br,
#                      'left2': Br, 'left3': Br} )


#-------------------------------------------------------------------------------                                 
# Evolve system through time
#-------------------------------------------------------------------------------

import time
t0 = time.time()
thisfile = project.integraltimeseries+'.csv'
fid = open(thisfile, 'w')
    
for t in domain.evolve(yieldstep = 10, finaltime = 600):
    domain.write_time()
    domain.write_boundary_statistics(tags = 'bottom')
    stagestep = domain.get_quantity('stage') 
    s = '%.2f, %.2f\n' %(t, stagestep.get_integral())
    fid.write(s)
    if allclose(t, 30):
        slump = Quantity(domain)
        slump.set_values(tsunami_source)
        domain.set_quantity('stage', slump + stagestep)

# save every two minutes leading up to interesting period
for t in domain.evolve(yieldstep = 120, finaltime = 660, # steps
                       skip_initial_step = True): 
    domain.write_time()
    domain.write_boundary_statistics(tags = 'bottom')
    # calculate integral
    thisstagestep = domain.get_quantity('stage') 
    s = '%.2f, %.2f\n' %(t, thisstagestep.get_integral())
    fid.write(s)
    
# save every thirty secs during interesting period
for t in domain.evolve(yieldstep = 30, finaltime = 3500, # steps
                       skip_initial_step = True):
    domain.write_time()
    domain.write_boundary_statistics(tags = 'bottom') #quantities = 'stage')
    # calculate integral
    thisstagestep = domain.get_quantity('stage') 
    s = '%.2f, %.2f\n' %(t, thisstagestep.get_integral())
    fid.write(s)

# save every two mins for next 5000 secs
for t in domain.evolve(yieldstep = 120, finaltime = 10000, # about 42 steps
                       skip_initial_step = True):
    domain.write_time()
    domain.write_boundary_statistics(tags = 'bottom') #quantities = 'stage')
    # calculate integral
    thisstagestep = domain.get_quantity('stage') 
    s = '%.2f, %.2f\n' %(t, thisstagestep.get_integral())
    fid.write(s)
    
# save every half hour to end of simulation    
for t in domain.evolve(yieldstep = 1800, finaltime = 10*60*60, # 14 steps
                       skip_initial_step = True):
    domain.write_time()
    domain.write_boundary_statistics(tags = 'bottom') #quantities = 'stage'
    # calculate integral
    thisstagestep = domain.get_quantity('stage') 
    s = '%.2f, %.2f\n' %(t, thisstagestep.get_integral())
    fid.write(s)

print 'That took %.2f seconds' %(time.time()-t0)