def initialize(self,
grid_size=(5, 5),
report_interval=5.0,
grid_orientation="vertical",
node_layout="rect",
show_plots=False,
cts_type="oriented_hex",
run_duration=1.0,
output_interval=1.0e99,
plot_every_transition=False,
**kwds):
# Remember the clock time, and calculate when we next want to report
# progress.
self.current_real_time = time.time()
self.next_report = self.current_real_time + report_interval
self.report_interval = report_interval
# Interval for output
self.output_interval = output_interval
# Duration for run
self.run_duration = run_duration
# Create a grid
self.create_grid_and_node_state_field(grid_size[0], grid_size[1],
grid_orientation, node_layout,
cts_type)
# Create the node-state dictionary
ns_dict = self.node_state_dictionary()
# Initialize values of the node-state grid
nsg = self.initialize_node_state_grid()
# Create the transition list
xn_list = self.transition_list()
# Create the CA object
if cts_type == "raster":
from landlab.ca.raster_cts import RasterCTS
self.ca = RasterCTS(self.grid, ns_dict, xn_list, nsg)
elif cts_type == "oriented_raster":
from landlab.ca.oriented_raster_cts import OrientedRasterCTS
self.ca = OrientedRasterCTS(self.grid, ns_dict, xn_list, nsg)
elif cts_type == "hex":
from landlab.ca.hex_cts import HexCTS
self.ca = HexCTS(self.grid, ns_dict, xn_list, nsg)
else:
from landlab.ca.oriented_hex_cts import OrientedHexCTS
self.ca = OrientedHexCTS(self.grid, ns_dict, xn_list, nsg)
# Initialize graphics
self._show_plots = show_plots
if show_plots:
self.initialize_plotting(**kwds)
def initialize(self, grid_size=(5, 5), report_interval=5.0,
grid_orientation='vertical', grid_shape='rect',
show_plots=False, cts_type='oriented_hex',
run_duration=1.0, output_interval=1.0e99,
plot_every_transition=False, **kwds):
# Remember the clock time, and calculate when we next want to report
# progress.
self.current_real_time = time.time()
self.next_report = self.current_real_time + report_interval
self.report_interval = report_interval
# Interval for output
self.output_interval = output_interval
# Duration for run
self.run_duration = run_duration
# Create a grid
self.create_grid_and_node_state_field(grid_size[0], grid_size[1],
grid_orientation, grid_shape,
cts_type)
# Create the node-state dictionary
ns_dict = self.node_state_dictionary()
# Initialize values of the node-state grid
nsg = self.initialize_node_state_grid()
# Create the transition list
xn_list = self.transition_list()
# Create the CA object
if cts_type == 'raster':
from landlab.ca.raster_cts import RasterCTS
self.ca = RasterCTS(self.grid, ns_dict, xn_list, nsg)
elif cts_type == 'oriented_raster':
from landlab.ca.oriented_raster_cts import OrientedRasterCTS
self.ca = OrientedRasterCTS(self.grid, ns_dict, xn_list, nsg)
elif cts_type == 'hex':
from landlab.ca.hex_cts import HexCTS
self.ca = HexCTS(self.grid, ns_dict, xn_list, nsg)
else:
from landlab.ca.oriented_hex_cts import OrientedHexCTS
self.ca = OrientedHexCTS(self.grid, ns_dict, xn_list, nsg)
# Initialize graphics
self._show_plots = show_plots
if show_plots == True:
self.initialize_plotting(**kwds)
def test_oriented_hex_cts():
"""Tests instantiation of an OrientedHexCTS() object"""
mg = HexModelGrid(3, 2, 1.0, orientation="vertical", reorient_links=True)
nsd = {0: "zero", 1: "one"}
xnlist = []
xnlist.append(Transition((0, 1, 0), (1, 1, 0), 1.0, "transitioning"))
nsg = mg.add_zeros("node", "node_state_grid")
ohcts = OrientedHexCTS(mg, nsd, xnlist, nsg)
assert ohcts.num_link_states == 12
# assert_array_equal(ohcts.link_orientation, [2, 1, 0, 0, 0, 2, 1, 0, 2, 1, 0])
assert_array_equal(ohcts.link_orientation, [2, 0, 1, 0, 2, 0, 1, 0, 2, 0, 1])
def test_oriented_hex_cts():
"""Tests instantiation of an OrientedHexCTS() object"""
mg = HexModelGrid(3, 2, 1.0, orientation='vertical', reorient_links=True)
nsd = {0 : 'zero', 1 : 'one'}
xnlist = []
xnlist.append(Transition((0,1,0), (1,1,0), 1.0, 'transitioning'))
nsg = mg.add_zeros('node', 'node_state_grid')
ohcts = OrientedHexCTS(mg, nsd, xnlist, nsg)
assert_equal(ohcts.num_link_states, 12)
#assert_array_equal(ohcts.link_orientation, [2, 1, 0, 0, 0, 2, 1, 0, 2, 1, 0])
assert_array_equal(ohcts.link_orientation, [2, 0, 1, 0, 2, 0, 1, 0, 2, 0, 1])
def test_shift_link_and_transition_data_upward():
"""Test the LatticeUplifter method that uplifts link data and tr'ns."""
mg = HexModelGrid((4, 3),
spacing=1.0,
orientation="vertical",
node_layout="rect")
nsd = {0: "yes", 1: "no"}
xnlist = []
xnlist.append(Transition((0, 0, 0), (1, 1, 0), 1.0, "frogging"))
xnlist.append(Transition((0, 0, 1), (1, 1, 1), 1.0, "frogging"))
xnlist.append(Transition((0, 0, 2), (1, 1, 2), 1.0, "frogging"))
nsg = mg.add_zeros("node_state_grid", at="node")
ohcts = OrientedHexCTS(mg, nsd, xnlist, nsg)
assert_array_equal(ohcts.link_state[mg.active_links],
[0, 4, 8, 8, 4, 0, 4, 8, 8, 4, 0])
assert_array_equal(ohcts.next_trn_id[mg.active_links],
[0, 1, 2, 2, 1, 0, 1, 2, 2, 1, 0])
assert_array_equal(
np.round(ohcts.next_update[mg.active_links], 2),
[0.8, 1.26, 0.92, 0.79, 0.55, 1.04, 0.58, 2.22, 3.31, 0.48, 1.57],
)
pq = ohcts.priority_queue
assert_equal(pq._queue[0][2], 19) # link for first event = 19, not shifted
assert_equal(round(pq._queue[0][0], 2), 0.48) # trn scheduled for t = 0.48
assert_equal(pq._queue[2][2], 14) # this event scheduled for link 15...
assert_equal(round(pq._queue[2][0], 2), 0.58) # ...trn sched for t = 0.58
lu = LatticeUplifter(grid=mg)
lu.shift_link_and_transition_data_upward(ohcts, 0.0)
# note new events lowest 5 links
assert_array_equal(
np.round(ohcts.next_update[mg.active_links], 2),
[0.75, 0.84, 2.6, 0.07, 0.09, 0.8, 0.02, 1.79, 1.51, 2.04, 3.85],
)
assert_equal(pq._queue[0][2], 14) # new soonest event
assert_equal(pq._queue[9][2], 13) # was previously 7, now shifted up...
assert_equal(round(pq._queue[9][0], 2), 0.8) # ...still sched for t = 0.80
class CTSModel(object):
"""
Implement a generic CellLab-CTS model.
This is the base class from which models should inherit.
"""
def __init__(self,
grid_size=(5, 5),
report_interval=5.0,
grid_orientation='vertical',
grid_shape='rect',
show_plots=False,
cts_type='oriented_hex',
run_duration=1.0,
output_interval=1.0e99,
plot_every_transition=False,
initial_state_grid=None,
prop_data=None,
prop_reset_value=None,
**kwds):
self.initialize(grid_size, report_interval, grid_orientation,
grid_shape, show_plots, cts_type, run_duration,
output_interval, plot_every_transition,
initial_state_grid, prop_data, prop_reset_value,
**kwds)
def initialize(self,
grid_size=(5, 5),
report_interval=5.0,
grid_orientation='vertical',
grid_shape='rect',
show_plots=False,
cts_type='oriented_hex',
run_duration=1.0,
output_interval=1.0e99,
plot_every_transition=False,
initial_state_grid=None,
prop_data=None,
prop_reset_value=None,
**kwds):
"""Initialize CTSModel."""
# Remember the clock time, and calculate when we next want to report
# progress.
self.current_real_time = time.time()
self.next_report = self.current_real_time + report_interval
self.report_interval = report_interval
# Interval for output
self.output_interval = output_interval
# Duration for run
self.run_duration = run_duration
# Create a grid
self.create_grid_and_node_state_field(grid_size[0], grid_size[1],
grid_orientation, grid_shape,
cts_type)
# If prop_data is a string, we assume it is a field name
if isinstance(prop_data, string_types):
prop_data = self.grid.add_zeros('node', prop_data)
# Create the node-state dictionary
ns_dict = self.node_state_dictionary()
# Initialize values of the node-state grid
if initial_state_grid is None:
nsg = self.initialize_node_state_grid()
else:
try:
nsg = initial_state_grid
self.grid.at_node['node_state'][:] = nsg
except:
#TODO: use new Messaging capability
print('If initial_state_grid given, must be array of int')
raise
# Create the transition list
xn_list = self.transition_list()
# Create the CA object
if cts_type == 'raster':
from landlab.ca.raster_cts import RasterCTS
self.ca = RasterCTS(self.grid, ns_dict, xn_list, nsg, prop_data,
prop_reset_value)
elif cts_type == 'oriented_raster':
from landlab.ca.oriented_raster_cts import OrientedRasterCTS
self.ca = OrientedRasterCTS(self.grid, ns_dict, xn_list, nsg,
prop_data, prop_reset_value)
elif cts_type == 'hex':
from landlab.ca.hex_cts import HexCTS
self.ca = HexCTS(self.grid, ns_dict, xn_list, nsg, prop_data,
prop_reset_value)
else:
from landlab.ca.oriented_hex_cts import OrientedHexCTS
self.ca = OrientedHexCTS(self.grid, ns_dict, xn_list, nsg,
prop_data, prop_reset_value)
# Initialize graphics
self._show_plots = show_plots
if show_plots == True:
self.initialize_plotting(**kwds)
def create_grid_and_node_state_field(self, num_rows, num_cols,
grid_orientation, grid_shape,
cts_type):
"""Create the grid and the field containing node states."""
if cts_type == 'raster' or cts_type == 'oriented_raster':
from landlab import RasterModelGrid
self.grid = RasterModelGrid(shape=(num_rows, num_cols),
spacing=1.0)
else:
from landlab import HexModelGrid
self.grid = HexModelGrid(num_rows,
num_cols,
1.0,
orientation=grid_orientation,
shape=grid_shape)
self.grid.add_zeros('node', 'node_state', dtype=int)
for edge in (self.grid.nodes_at_right_edge,
self.grid.nodes_at_top_edge):
self.grid.status_at_node[edge] = CLOSED_BOUNDARY
def node_state_dictionary(self):
"""Create and return a dictionary of all possible node (cell) states.
This method creates a default set of states (just two); it is a
template meant to be overridden.
"""
ns_dict = {0: 'on', 1: 'off'}
return ns_dict
def transition_list(self):
"""Create and return a list of transition objects.
This method creates a default set of transitions (just two); it is a
template meant to be overridden.
"""
xn_list = []
xn_list.append(Transition((0, 1, 0), (1, 0, 0), 1.0))
xn_list.append(Transition((1, 0, 0), (0, 1, 0), 1.0))
return xn_list
def write_output(self, grid, outfilename, iteration):
"""Write output to file (currently netCDF)."""
filename = outfilename + str(iteration).zfill(4) + '.nc'
save_grid(grid, filename)
def initialize_node_state_grid(self):
"""Initialize values in the node-state grid.
This method should be overridden. The default is random "on" and "off".
"""
num_states = 2
for i in range(self.grid.number_of_nodes):
self.grid.at_node['node_state'][i] = random.randint(num_states)
return self.grid.at_node['node_state']
def initialize_plotting(self, **kwds):
"""Create and configure CAPlotter object."""
self.ca_plotter = CAPlotter(self.ca, **kwds)
self.ca_plotter.update_plot()
axis('off')
def run_for(self, dt):
self.ca.run(self.ca.current_time + dt, self.ca.node_state)
def main():
# INITIALIZE
# User-defined parameters
nr = 41
nc = 61
g = 1.0
f = 0.7
silo_y0 = 30.0
silo_opening_half_width = 6
plot_interval = 10.0
run_duration = 240.0
report_interval = 300.0 # report interval, in real-time seconds
p_init = 0.4 # probability that a cell is occupied at start
plot_every_transition = False
# Remember the clock time, and calculate when we next want to report
# progress.
current_real_time = time.time()
next_report = current_real_time + report_interval
# Create a grid
hmg = HexModelGrid(
nr, nc, 1.0, orientation="vertical", shape="rect", reorient_links=True
)
# Set up the states and pair transitions.
# Transition data here represent particles moving on a lattice: one state
# per direction (for 6 directions), plus an empty state, a stationary
# state, and a wall state.
ns_dict = {
0: "empty",
1: "moving up",
2: "moving right and up",
3: "moving right and down",
4: "moving down",
5: "moving left and down",
6: "moving left and up",
7: "rest",
8: "wall",
}
xn_list = setup_transition_list(g, f)
# Create data and initialize values.
node_state_grid = hmg.add_zeros("node", "node_state_grid")
# Make the grid boundary all wall particles
node_state_grid[hmg.boundary_nodes] = 8
# Place wall particles to form the base of the silo, initially closed
tan30deg = numpy.tan(numpy.pi / 6.)
rampy1 = silo_y0 - hmg.node_x * tan30deg
rampy2 = silo_y0 - ((nc * 0.866 - 1.) - hmg.node_x) * tan30deg
rampy = numpy.maximum(rampy1, rampy2)
(ramp_nodes,) = numpy.where(
numpy.logical_and(hmg.node_y > rampy - 0.5, hmg.node_y < rampy + 0.5)
)
node_state_grid[ramp_nodes] = 8
# Seed the grid interior with randomly oriented particles
for i in hmg.core_nodes:
if hmg.node_y[i] > rampy[i] and random.random() < p_init:
node_state_grid[i] = random.randint(1, 7)
# Create the CA model
ca = OrientedHexCTS(hmg, ns_dict, xn_list, node_state_grid)
import matplotlib
rock = (0.0, 0.0, 0.0) # '#5F594D'
sed = (0.6, 0.6, 0.6) # '#A4874B'
# sky = '#CBD5E1'
# sky = '#85A5CC'
sky = (1.0, 1.0, 1.0) # '#D0E4F2'
mob = (0.3, 0.3, 0.3) # '#D98859'
# mob = '#DB764F'
# mob = '#FFFF00'
# sed = '#CAAE98'
# clist = [(0.5, 0.9, 0.9),mob, mob, mob, mob, mob, mob,'#CD6839',(0.3,0.3,0.3)]
clist = [sky, mob, mob, mob, mob, mob, mob, sed, rock]
my_cmap = matplotlib.colors.ListedColormap(clist)
# Create a CAPlotter object for handling screen display
ca_plotter = CAPlotter(ca, cmap=my_cmap)
k = 0
# Plot the initial grid
ca_plotter.update_plot()
# RUN
# Run with closed silo
current_time = 0.0
while current_time < run_duration:
# Once in a while, print out simulation and real time to let the user
# know that the sim is running ok
current_real_time = time.time()
if current_real_time >= next_report:
print(
"Current sim time",
current_time,
"(",
100 * current_time / run_duration,
"%)",
)
next_report = current_real_time + report_interval
# Run the model forward in time until the next output step
ca.run(
current_time + plot_interval,
ca.node_state,
plot_each_transition=plot_every_transition,
plotter=ca_plotter,
)
current_time += plot_interval
# Plot the current grid
ca_plotter.update_plot()
# Open the silo
xmid = nc * 0.866 * 0.5
for i in range(hmg.number_of_nodes):
if (
node_state_grid[i] == 8
and hmg.node_x[i] > (xmid - silo_opening_half_width)
and hmg.node_x[i] < (xmid + silo_opening_half_width)
and hmg.node_y[i] > 0
and hmg.node_y[i] < 38.0
):
node_state_grid[i] = 0
# Create the CA model
ca = OrientedHexCTS(hmg, ns_dict, xn_list, node_state_grid)
# Create a CAPlotter object for handling screen display
ca_plotter = CAPlotter(ca)
# Plot the initial grid
ca_plotter.update_plot()
# Re-run with open silo
savefig("silo" + str(k) + ".png")
k += 1
current_time = 0.0
while current_time < 5 * run_duration:
# Once in a while, print out simulation and real time to let the user
# know that the sim is running ok
current_real_time = time.time()
if current_real_time >= next_report:
print(
"Current sim time",
current_time,
"(",
100 * current_time / run_duration,
"%)",
)
next_report = current_real_time + report_interval
# Run the model forward in time until the next output step
ca.run(
current_time + plot_interval,
ca.node_state,
plot_each_transition=plot_every_transition,
plotter=ca_plotter,
)
current_time += plot_interval
# Plot the current grid
ca_plotter.update_plot()
savefig("silo" + str(k) + ".png")
k += 1
# FINALIZE
# Plot
ca_plotter.finalize()
def main():
# INITIALIZE
# User-defined parameters
nr = 41
nc = 61
g = 1.0
f = 0.7
silo_y0 = 30.0
silo_opening_half_width = 6
plot_interval = 10.0
run_duration = 240.0
report_interval = 300.0 # report interval, in real-time seconds
p_init = 0.4 # probability that a cell is occupied at start
plot_every_transition = False
# Remember the clock time, and calculate when we next want to report
# progress.
current_real_time = time.time()
next_report = current_real_time + report_interval
# Create a grid
hmg = HexModelGrid(nr,
nc,
1.0,
orientation='vertical',
shape='rect',
reorient_links=True)
# Set up the states and pair transitions.
# Transition data here represent particles moving on a lattice: one state
# per direction (for 6 directions), plus an empty state, a stationary
# state, and a wall state.
ns_dict = {
0: 'empty',
1: 'moving up',
2: 'moving right and up',
3: 'moving right and down',
4: 'moving down',
5: 'moving left and down',
6: 'moving left and up',
7: 'rest',
8: 'wall'
}
xn_list = setup_transition_list(g, f)
# Create data and initialize values.
node_state_grid = hmg.add_zeros('node', 'node_state_grid')
# Make the grid boundary all wall particles
node_state_grid[hmg.boundary_nodes] = 8
# Place wall particles to form the base of the silo, initially closed
tan30deg = numpy.tan(numpy.pi / 6.)
rampy1 = silo_y0 - hmg.node_x * tan30deg
rampy2 = silo_y0 - ((nc * 0.866 - 1.) - hmg.node_x) * tan30deg
rampy = numpy.maximum(rampy1, rampy2)
(ramp_nodes, ) = numpy.where(numpy.logical_and(hmg.node_y>rampy-0.5, \
hmg.node_y<rampy+0.5))
node_state_grid[ramp_nodes] = 8
# Seed the grid interior with randomly oriented particles
for i in hmg.core_nodes:
if hmg.node_y[i] > rampy[i] and random.random() < p_init:
node_state_grid[i] = random.randint(1, 7)
# Create the CA model
ca = OrientedHexCTS(hmg, ns_dict, xn_list, node_state_grid)
import matplotlib
rock = (0.0, 0.0, 0.0) #'#5F594D'
sed = (0.6, 0.6, 0.6) #'#A4874B'
#sky = '#CBD5E1'
#sky = '#85A5CC'
sky = (1.0, 1.0, 1.0) #'#D0E4F2'
mob = (0.3, 0.3, 0.3) #'#D98859'
#mob = '#DB764F'
#mob = '#FFFF00'
#sed = '#CAAE98'
#clist = [(0.5, 0.9, 0.9),mob, mob, mob, mob, mob, mob,'#CD6839',(0.3,0.3,0.3)]
clist = [sky, mob, mob, mob, mob, mob, mob, sed, rock]
my_cmap = matplotlib.colors.ListedColormap(clist)
# Create a CAPlotter object for handling screen display
ca_plotter = CAPlotter(ca, cmap=my_cmap)
k = 0
# Plot the initial grid
ca_plotter.update_plot()
# RUN
# Run with closed silo
current_time = 0.0
while current_time < run_duration:
# Once in a while, print out simulation and real time to let the user
# know that the sim is running ok
current_real_time = time.time()
if current_real_time >= next_report:
print 'Current sim time', current_time, '(', 100 * current_time / run_duration, '%)'
next_report = current_real_time + report_interval
# Run the model forward in time until the next output step
ca.run(current_time + plot_interval,
ca.node_state,
plot_each_transition=plot_every_transition,
plotter=ca_plotter)
current_time += plot_interval
# Plot the current grid
ca_plotter.update_plot()
# Open the silo
xmid = nc * 0.866 * 0.5
for i in range(hmg.number_of_nodes):
if node_state_grid[i]==8 and hmg.node_x[i]>(xmid-silo_opening_half_width) \
and hmg.node_x[i]<(xmid+silo_opening_half_width) \
and hmg.node_y[i]>0 and hmg.node_y[i]<38.0:
node_state_grid[i] = 0
# Create the CA model
ca = OrientedHexCTS(hmg, ns_dict, xn_list, node_state_grid)
# Create a CAPlotter object for handling screen display
ca_plotter = CAPlotter(ca)
# Plot the initial grid
ca_plotter.update_plot()
# Re-run with open silo
savefig('silo' + str(k) + '.png')
k += 1
current_time = 0.0
while current_time < 5 * run_duration:
# Once in a while, print out simulation and real time to let the user
# know that the sim is running ok
current_real_time = time.time()
if current_real_time >= next_report:
print 'Current sim time', current_time, '(', 100 * current_time / run_duration, '%)'
next_report = current_real_time + report_interval
# Run the model forward in time until the next output step
ca.run(current_time + plot_interval,
ca.node_state,
plot_each_transition=plot_every_transition,
plotter=ca_plotter)
current_time += plot_interval
# Plot the current grid
ca_plotter.update_plot()
savefig('silo' + str(k) + '.png')
k += 1
# FINALIZE
# Plot
ca_plotter.finalize()
def main():
# INITIALIZE
# User-defined parameters
nr = 21
nc = 21
plot_interval = 0.5
run_duration = 25.0
report_interval = 5.0 # report interval, in real-time seconds
# Remember the clock time, and calculate when we next want to report
# progress.
current_real_time = time.time()
next_report = current_real_time + report_interval
# Create a grid
hmg = HexModelGrid(nr, nc, 1.0, orientation='vertical', reorient_links=True)
# Close the grid boundaries
hmg.set_closed_nodes(hmg.open_boundary_nodes)
# Set up the states and pair transitions.
# Transition data here represent the disease status of a population.
ns_dict = { 0 : 'fluid', 1 : 'grain' }
xn_list = setup_transition_list()
# Create data and initialize values. We start with the 3 middle columns full
# of grains, and the others empty.
node_state_grid = hmg.add_zeros('node', 'node_state_grid')
middle = 0.25*(nc-1)*sqrt(3)
is_middle_cols = logical_and(hmg.node_x<middle+1., hmg.node_x>middle-1.)
node_state_grid[where(is_middle_cols)[0]] = 1
# Create the CA model
ca = OrientedHexCTS(hmg, ns_dict, xn_list, node_state_grid)
# Create a CAPlotter object for handling screen display
ca_plotter = CAPlotter(ca)
# Plot the initial grid
ca_plotter.update_plot()
# RUN
current_time = 0.0
while current_time < run_duration:
# Once in a while, print out simulation and real time to let the user
# know that the sim is running ok
current_real_time = time.time()
if current_real_time >= next_report:
print('Current sim time',current_time,'(',100*current_time/run_duration,'%)')
next_report = current_real_time + report_interval
# Run the model forward in time until the next output step
ca.run(current_time+plot_interval, ca.node_state,
plot_each_transition=False)
current_time += plot_interval
# Plot the current grid
ca_plotter.update_plot()
# FINALIZE
# Plot
ca_plotter.finalize()
def initialize(self, grid_size=(5, 5), report_interval=5.0,
grid_orientation='vertical', grid_shape='rect',
show_plots=False, cts_type='oriented_hex',
run_duration=1.0, output_interval=1.0e99,
plot_every_transition=False, initial_state_grid=None,
prop_data=None, prop_reset_value=None,
closed_boundaries=(False, False, False, False), **kwds):
"""Initialize CTSModel."""
# Remember the clock time, and calculate when we next want to report
# progress.
self.current_real_time = time.time()
self.next_report = self.current_real_time + report_interval
self.report_interval = report_interval
# Interval for output
self.output_interval = output_interval
# Duration for run
self.run_duration = run_duration
# Create a grid
self.create_grid_and_node_state_field(grid_size[0], grid_size[1],
grid_orientation, grid_shape,
cts_type, closed_boundaries)
# If prop_data is a string, we assume it is a field name
if isinstance(prop_data, string_types):
prop_data = self.grid.add_zeros('node', prop_data)
# Create the node-state dictionary
ns_dict = self.node_state_dictionary()
# Initialize values of the node-state grid
if initial_state_grid is None:
nsg = self.initialize_node_state_grid()
else:
try:
nsg = initial_state_grid
self.grid.at_node['node_state'][:] = nsg
except:
#TODO: use new Messaging capability
print('If initial_state_grid given, must be array of int')
raise
# Create the transition list
xn_list = self.transition_list()
# Create the CA object
if cts_type == 'raster':
from landlab.ca.raster_cts import RasterCTS
self.ca = RasterCTS(self.grid, ns_dict, xn_list, nsg, prop_data,
prop_reset_value)
elif cts_type == 'oriented_raster':
from landlab.ca.oriented_raster_cts import OrientedRasterCTS
self.ca = OrientedRasterCTS(self.grid, ns_dict, xn_list, nsg,
prop_data, prop_reset_value)
elif cts_type == 'hex':
from landlab.ca.hex_cts import HexCTS
self.ca = HexCTS(self.grid, ns_dict, xn_list, nsg, prop_data,
prop_reset_value)
else:
from landlab.ca.oriented_hex_cts import OrientedHexCTS
self.ca = OrientedHexCTS(self.grid, ns_dict, xn_list, nsg,
prop_data, prop_reset_value)
# Initialize graphics
self._show_plots = show_plots
if show_plots == True:
self.initialize_plotting(**kwds)
class CTSModel(object):
"""
Implement a generic CellLab-CTS model.
This is the base class from which models should inherit.
"""
def __init__(self, grid_size=(5, 5), report_interval=5.0,
grid_orientation='vertical', grid_shape='rect',
show_plots=False, cts_type='oriented_hex',
run_duration=1.0, output_interval=1.0e99,
plot_every_transition=False, initial_state_grid=None,
prop_data=None, prop_reset_value=None,
closed_boundaries=(False, False, False, False), **kwds):
self.initialize(grid_size, report_interval, grid_orientation,
grid_shape, show_plots, cts_type, run_duration,
output_interval, plot_every_transition,
initial_state_grid, prop_data, prop_reset_value,
closed_boundaries, **kwds)
def initialize(self, grid_size=(5, 5), report_interval=5.0,
grid_orientation='vertical', grid_shape='rect',
show_plots=False, cts_type='oriented_hex',
run_duration=1.0, output_interval=1.0e99,
plot_every_transition=False, initial_state_grid=None,
prop_data=None, prop_reset_value=None,
closed_boundaries=(False, False, False, False), **kwds):
"""Initialize CTSModel."""
# Remember the clock time, and calculate when we next want to report
# progress.
self.current_real_time = time.time()
self.next_report = self.current_real_time + report_interval
self.report_interval = report_interval
# Interval for output
self.output_interval = output_interval
# Duration for run
self.run_duration = run_duration
# Create a grid
self.create_grid_and_node_state_field(grid_size[0], grid_size[1],
grid_orientation, grid_shape,
cts_type, closed_boundaries)
# If prop_data is a string, we assume it is a field name
if isinstance(prop_data, string_types):
prop_data = self.grid.add_zeros('node', prop_data)
# Create the node-state dictionary
ns_dict = self.node_state_dictionary()
# Initialize values of the node-state grid
if initial_state_grid is None:
nsg = self.initialize_node_state_grid()
else:
try:
nsg = initial_state_grid
self.grid.at_node['node_state'][:] = nsg
except:
#TODO: use new Messaging capability
print('If initial_state_grid given, must be array of int')
raise
# Create the transition list
xn_list = self.transition_list()
# Create the CA object
if cts_type == 'raster':
from landlab.ca.raster_cts import RasterCTS
self.ca = RasterCTS(self.grid, ns_dict, xn_list, nsg, prop_data,
prop_reset_value)
elif cts_type == 'oriented_raster':
from landlab.ca.oriented_raster_cts import OrientedRasterCTS
self.ca = OrientedRasterCTS(self.grid, ns_dict, xn_list, nsg,
prop_data, prop_reset_value)
elif cts_type == 'hex':
from landlab.ca.hex_cts import HexCTS
self.ca = HexCTS(self.grid, ns_dict, xn_list, nsg, prop_data,
prop_reset_value)
else:
from landlab.ca.oriented_hex_cts import OrientedHexCTS
self.ca = OrientedHexCTS(self.grid, ns_dict, xn_list, nsg,
prop_data, prop_reset_value)
# Initialize graphics
self._show_plots = show_plots
if show_plots == True:
self.initialize_plotting(**kwds)
def _set_closed_boundaries_for_hex_grid(self, closed_boundaries):
"""Setup one or more closed boundaries for a hex grid.
Parameters
----------
closed_boundaries : 4-element tuple of bool\
Whether right, top, left, and bottom edges have closed nodes
Examples
--------
>>> from grainhill import CTSModel
>>> cm = CTSModel(closed_boundaries=(True, True, True, True))
>>> cm.grid.status_at_node
array([4, 4, 4, 4, 4, 4, 0, 4, 0, 0, 4, 0, 4, 0, 0, 4, 0, 4, 0, 0, 4, 4,
4, 4, 4], dtype=uint8)
"""
g = self.grid
if closed_boundaries[0]:
g.status_at_node[g.nodes_at_right_edge] = CLOSED_BOUNDARY
if closed_boundaries[1]:
g.status_at_node[g.nodes_at_top_edge] = CLOSED_BOUNDARY
if closed_boundaries[2]:
g.status_at_node[g.nodes_at_left_edge] = CLOSED_BOUNDARY
if closed_boundaries[3]:
g.status_at_node[g.nodes_at_bottom_edge] = CLOSED_BOUNDARY
def create_grid_and_node_state_field(self, num_rows, num_cols,
grid_orientation, grid_shape,
cts_type, closed_bounds):
"""Create the grid and the field containing node states."""
if cts_type == 'raster' or cts_type == 'oriented_raster':
from landlab import RasterModelGrid
self.grid = RasterModelGrid(shape=(num_rows, num_cols),
spacing=1.0)
self.grid.set_closed_boundaries_at_grid_edges(closed_bounds[0],
closed_bounds[1],
closed_bounds[2],
closed_bounds[3])
else:
from landlab import HexModelGrid
self.grid = HexModelGrid(num_rows, num_cols, 1.0,
orientation=grid_orientation,
shape=grid_shape)
if True in closed_bounds:
self._set_closed_boundaries_for_hex_grid(closed_bounds)
self.grid.add_zeros('node', 'node_state', dtype=int)
def node_state_dictionary(self):
"""Create and return a dictionary of all possible node (cell) states.
This method creates a default set of states (just two); it is a
template meant to be overridden.
"""
ns_dict = { 0 : 'on',
1 : 'off'}
return ns_dict
def transition_list(self):
"""Create and return a list of transition objects.
This method creates a default set of transitions (just two); it is a
template meant to be overridden.
"""
xn_list = []
xn_list.append(Transition((0, 1, 0), (1, 0, 0), 1.0))
xn_list.append(Transition((1, 0, 0), (0, 1, 0), 1.0))
return xn_list
def write_output(self, grid, outfilename, iteration):
"""Write output to file (currently netCDF)."""
filename = outfilename + str(iteration).zfill(4) + '.nc'
save_grid(grid, filename)
def initialize_node_state_grid(self):
"""Initialize values in the node-state grid.
This method should be overridden. The default is random "on" and "off".
"""
num_states = 2
for i in range(self.grid.number_of_nodes):
self.grid.at_node['node_state'][i] = random.randint(num_states)
return self.grid.at_node['node_state']
def initialize_plotting(self, **kwds):
"""Create and configure CAPlotter object."""
self.ca_plotter = CAPlotter(self.ca, **kwds)
self.ca_plotter.update_plot()
axis('off')
def run_for(self, dt):
self.ca.run(self.ca.current_time + dt, self.ca.node_state)
class CTSModel(object):
"""
Implement a generic CellLab-CTS model.
This is the base class from which models should inherit.
"""
def __init__(self, grid_size=(5, 5), report_interval=5.0,
grid_orientation='vertical', grid_shape='rect',
show_plots=False, cts_type='oriented_hex',
run_duration=1.0, output_interval=1.0e99,
plot_every_transition=False, **kwds):
self.initialize(grid_size, report_interval, grid_orientation,
grid_shape, show_plots, cts_type, run_duration,
output_interval, plot_every_transition, **kwds)
def initialize(self, grid_size=(5, 5), report_interval=5.0,
grid_orientation='vertical', grid_shape='rect',
show_plots=False, cts_type='oriented_hex',
run_duration=1.0, output_interval=1.0e99,
plot_every_transition=False, **kwds):
# Remember the clock time, and calculate when we next want to report
# progress.
self.current_real_time = time.time()
self.next_report = self.current_real_time + report_interval
self.report_interval = report_interval
# Interval for output
self.output_interval = output_interval
# Duration for run
self.run_duration = run_duration
# Create a grid
self.create_grid_and_node_state_field(grid_size[0], grid_size[1],
grid_orientation, grid_shape,
cts_type)
# Create the node-state dictionary
ns_dict = self.node_state_dictionary()
# Initialize values of the node-state grid
nsg = self.initialize_node_state_grid()
# Create the transition list
xn_list = self.transition_list()
# Create the CA object
if cts_type == 'raster':
from landlab.ca.raster_cts import RasterCTS
self.ca = RasterCTS(self.grid, ns_dict, xn_list, nsg)
elif cts_type == 'oriented_raster':
from landlab.ca.oriented_raster_cts import OrientedRasterCTS
self.ca = OrientedRasterCTS(self.grid, ns_dict, xn_list, nsg)
elif cts_type == 'hex':
from landlab.ca.hex_cts import HexCTS
self.ca = HexCTS(self.grid, ns_dict, xn_list, nsg)
else:
from landlab.ca.oriented_hex_cts import OrientedHexCTS
self.ca = OrientedHexCTS(self.grid, ns_dict, xn_list, nsg)
# Initialize graphics
self._show_plots = show_plots
if show_plots == True:
self.initialize_plotting(**kwds)
def create_grid_and_node_state_field(self, num_rows, num_cols,
grid_orientation, grid_shape,
cts_type):
"""Create the grid and the field containing node states."""
if cts_type == 'raster' or cts_type == 'oriented_raster':
from landlab import RasterModelGrid
self.grid = RasterModelGrid(shape=(num_rows, num_cols),
spacing=1.0)
else:
from landlab import HexModelGrid
self.grid = HexModelGrid(num_rows, num_cols, 1.0,
orientation=grid_orientation,
shape=grid_shape)
self.grid.add_zeros('node', 'node_state', dtype=int)
def node_state_dictionary(self):
"""Create and return a dictionary of all possible node (cell) states.
This method creates a default set of states (just two); it is a
template meant to be overridden.
"""
ns_dict = { 0 : 'on',
1 : 'off'}
return ns_dict
def transition_list(self):
"""Create and return a list of transition objects.
This method creates a default set of transitions (just two); it is a
template meant to be overridden.
"""
xn_list = []
xn_list.append(Transition((0, 1, 0), (1, 0, 0), 1.0))
xn_list.append(Transition((1, 0, 0), (0, 1, 0), 1.0))
return xn_list
def write_output(self, grid, outfilename, iteration):
"""Write output to file (currently netCDF)."""
filename = outfilename + str(iteration).zfill(4) + '.nc'
save_grid(grid, filename)
def initialize_node_state_grid(self):
"""Initialize values in the node-state grid.
This method should be overridden. The default is random "on" and "off".
"""
num_states = 2
for i in range(self.grid.number_of_nodes):
self.grid.at_node['node_state'][i] = random.randint(num_states)
return self.grid.at_node['node_state']
def initialize_plotting(self, **kwds):
"""Create and configure CAPlotter object."""
self.ca_plotter = CAPlotter(self.ca, **kwds)
self.ca_plotter.update_plot()
axis('off')
def run_for(self, dt):
self.ca.run(self.ca.current_time + dt, self.ca.node_state)
def main():
# INITIALIZE
# User-defined parameters
nr = 41
nc = 61
g = 0.8
f = 1.0
plot_interval = 1.0
run_duration = 200.0
report_interval = 5.0 # report interval, in real-time seconds
p_init = 0.4 # probability that a cell is occupied at start
plot_every_transition = False
# Remember the clock time, and calculate when we next want to report
# progress.
current_real_time = time.time()
next_report = current_real_time + report_interval
# Create a grid
hmg = HexModelGrid(nr, nc, 1.0, orientation='vertical', reorient_links=True)
# Close the grid boundaries
#hmg.set_closed_nodes(hmg.open_boundary_nodes)
# Set up the states and pair transitions.
# Transition data here represent particles moving on a lattice: one state
# per direction (for 6 directions), plus an empty state, a stationary
# state, and a wall state.
ns_dict = { 0 : 'empty',
1 : 'moving up',
2 : 'moving right and up',
3 : 'moving right and down',
4 : 'moving down',
5 : 'moving left and down',
6 : 'moving left and up',
7 : 'rest',
8 : 'wall'}
xn_list = setup_transition_list(g, f)
# Create data and initialize values.
node_state_grid = hmg.add_zeros('node', 'node_state_grid')
# Make the grid boundary all wall particles
node_state_grid[hmg.boundary_nodes] = 8
# Seed the grid interior with randomly oriented particles
for i in hmg.core_nodes:
if random.random()<p_init:
node_state_grid[i] = random.randint(1, 7)
# Create the CA model
ca = OrientedHexCTS(hmg, ns_dict, xn_list, node_state_grid)
# Create a CAPlotter object for handling screen display
ca_plotter = CAPlotter(ca)
# Plot the initial grid
ca_plotter.update_plot()
# RUN
current_time = 0.0
while current_time < run_duration:
# Once in a while, print out simulation and real time to let the user
# know that the sim is running ok
current_real_time = time.time()
if current_real_time >= next_report:
print('Current sim time',current_time,'(',100*current_time/run_duration,'%)')
next_report = current_real_time + report_interval
# Run the model forward in time until the next output step
ca.run(current_time+plot_interval, ca.node_state,
plot_each_transition=plot_every_transition, plotter=ca_plotter)
current_time += plot_interval
# Plot the current grid
ca_plotter.update_plot()
# FINALIZE
# Plot
ca_plotter.finalize()
def run(uplift_interval, d): #d_ratio_exp):
# INITIALIZE
#uplift_interval = 1e7
#filenm = 'test_output'
#imagenm = 'Hill141213/hill'+str(int(d_ratio_exp))+'d'
# Remember the clock time, and calculate when we next want to report
# progress.
current_real_time = time.time()
next_report = current_real_time + report_interval
next_uplift = uplift_interval
# Create a grid
hmg = HexModelGrid(nr,
nc,
1.0,
orientation='vertical',
shape='rect',
reorient_links=True)
# Close the right-hand grid boundaries
#hmg.set_closed_nodes(arange((nc-1)*nr, hmg.number_of_nodes))
# Set up the states and pair transitions.
# Transition data here represent particles moving on a lattice: one state
# per direction (for 6 directions), plus an empty state, a stationary
# state, and a wall state.
ns_dict = {
0: 'empty',
1: 'moving up',
2: 'moving right and up',
3: 'moving right and down',
4: 'moving down',
5: 'moving left and down',
6: 'moving left and up',
7: 'rest',
8: 'wall'
}
xn_list = setup_transition_list(g, f, d)
#xn_list = []
#xn_list.append( Transition((0,1,0), (0,7,0), g, 'gravity 1') )
# Create data and initialize values.
node_state_grid = hmg.add_zeros('node', 'node_state_grid', dtype=int)
# Lower rows get resting particles
if nc % 2 == 0: # if even num cols, bottom right is...
bottom_right = nc - 1
else:
bottom_right = nc // 2
right_side_x = 0.866025403784 * (nc - 1)
for i in range(hmg.number_of_nodes):
if hmg.node_y[i] < 3.0:
if hmg.node_x[i] > 0.0 and hmg.node_x[i] < right_side_x:
node_state_grid[i] = 7
#elif hmg.node_x[i]>((nc-1)*0.866):
# node_state_grid[i] = 8
node_state_grid[0] = 8 # bottom left
node_state_grid[bottom_right] = 8
#for i in range(hmg.number_of_nodes):
# print i, hmg.node_x[i], hmg.node_y[i], node_state_grid[i]
# Create an uplift object
uplifter = LatticeUplifter(hmg, node_state_grid)
# Create the CA model
ca = OrientedHexCTS(hmg, ns_dict, xn_list, node_state_grid)
# Create a CAPlotter object for handling screen display
# potential colors: red3='#CD0000'
#mob = 'r'
#rock = '#5F594D'
sed = '#A4874B'
#sky = '#CBD5E1'
#sky = '#85A5CC'
sky = '#D0E4F2'
rock = '#000000' #sky
mob = '#D98859'
#mob = '#DB764F'
#mob = '#FFFF00'
#sed = '#CAAE98'
#clist = [(0.5, 0.9, 0.9),mob, mob, mob, mob, mob, mob,'#CD6839',(0.3,0.3,0.3)]
clist = [sky, mob, mob, mob, mob, mob, mob, sed, rock]
my_cmap = matplotlib.colors.ListedColormap(clist)
ca_plotter = CAPlotter(ca, cmap=my_cmap)
k = 0
# Plot the initial grid
ca_plotter.update_plot()
axis('off')
#savefig(imagenm+str(k)+'.png')
k += 1
# Write output for initial grid
#write_output(hmg, filenm, 0)
#output_iteration = 1
# Create an array to store the numbers of states at each plot interval
#nstates = zeros((9, int(run_duration/plot_interval)))
#k = 0
# Work out the next times to plot and output
next_output = output_interval
next_plot = plot_interval
# RUN
current_time = 0.0
while current_time < run_duration:
# Figure out what time to run to this iteration
next_pause = min(next_output, next_plot)
next_pause = min(next_pause, next_uplift)
next_pause = min(next_pause, run_duration)
# Once in a while, print out simulation and real time to let the user
# know that the sim is running ok
current_real_time = time.time()
if current_real_time >= next_report:
print 'Current sim time', current_time, '(', 100 * current_time / run_duration, '%)'
next_report = current_real_time + report_interval
# Run the model forward in time until the next output step
print('Running to...' + str(next_pause))
ca.run(next_pause, ca.node_state) #,
#plot_each_transition=plot_every_transition, plotter=ca_plotter)
current_time = next_pause
# Handle output to file
if current_time >= next_output:
#write_output(hmg, filenm, output_iteration)
#output_iteration += 1
next_output += output_interval
# Handle plotting on display
if current_time >= next_plot:
#node_state_grid[hmg.number_of_node_rows-1] = 8
ca_plotter.update_plot()
axis('off')
next_plot += plot_interval
# Handle uplift
if current_time >= next_uplift:
uplifter.uplift_interior_nodes(rock_state=7)
ca.update_link_states_and_transitions(current_time)
next_uplift += uplift_interval
print('Finished with main loop')
def initialize(self,
grid_size=(5, 5),
report_interval=5.0,
grid_orientation='vertical',
grid_shape='rect',
show_plots=False,
cts_type='oriented_hex',
run_duration=1.0,
output_interval=1.0e99,
plot_every_transition=False,
initial_state_grid=None,
prop_data=None,
prop_reset_value=None,
**kwds):
"""Initialize CTSModel."""
# Remember the clock time, and calculate when we next want to report
# progress.
self.current_real_time = time.time()
self.next_report = self.current_real_time + report_interval
self.report_interval = report_interval
# Interval for output
self.output_interval = output_interval
# Duration for run
self.run_duration = run_duration
# Create a grid
self.create_grid_and_node_state_field(grid_size[0], grid_size[1],
grid_orientation, grid_shape,
cts_type)
# If prop_data is a string, we assume it is a field name
if isinstance(prop_data, string_types):
prop_data = self.grid.add_zeros('node', prop_data)
# Create the node-state dictionary
ns_dict = self.node_state_dictionary()
# Initialize values of the node-state grid
if initial_state_grid is None:
nsg = self.initialize_node_state_grid()
else:
try:
nsg = initial_state_grid
self.grid.at_node['node_state'][:] = nsg
except:
#TODO: use new Messaging capability
print('If initial_state_grid given, must be array of int')
raise
# Create the transition list
xn_list = self.transition_list()
# Create the CA object
if cts_type == 'raster':
from landlab.ca.raster_cts import RasterCTS
self.ca = RasterCTS(self.grid, ns_dict, xn_list, nsg, prop_data,
prop_reset_value)
elif cts_type == 'oriented_raster':
from landlab.ca.oriented_raster_cts import OrientedRasterCTS
self.ca = OrientedRasterCTS(self.grid, ns_dict, xn_list, nsg,
prop_data, prop_reset_value)
elif cts_type == 'hex':
from landlab.ca.hex_cts import HexCTS
self.ca = HexCTS(self.grid, ns_dict, xn_list, nsg, prop_data,
prop_reset_value)
else:
from landlab.ca.oriented_hex_cts import OrientedHexCTS
self.ca = OrientedHexCTS(self.grid, ns_dict, xn_list, nsg,
prop_data, prop_reset_value)
# Initialize graphics
self._show_plots = show_plots
if show_plots == True:
self.initialize_plotting(**kwds)
def main():
# INITIALIZE
# User-defined parameters
nr = 41
nc = 61
g = 0.05
plot_interval = 1.0
run_duration = 100.0
report_interval = 5.0 # report interval, in real-time seconds
p_init = 0.1 # probability that a cell is occupied at start
plot_every_transition = False
# Remember the clock time, and calculate when we next want to report
# progress.
current_real_time = time.time()
next_report = current_real_time + report_interval
# Create a grid
hmg = HexModelGrid(nr,
nc,
1.0,
orientation='vertical',
reorient_links=True)
# Close the grid boundaries
#hmg.set_closed_nodes(hmg.open_boundary_nodes)
# Set up the states and pair transitions.
# Transition data here represent particles moving on a lattice: one state
# per direction (for 6 directions), plus an empty state, a stationary
# state, and a wall state.
ns_dict = {
0: 'empty',
1: 'moving up',
2: 'moving right and up',
3: 'moving right and down',
4: 'moving down',
5: 'moving left and down',
6: 'moving left and up',
7: 'rest',
8: 'wall'
}
xn_list = setup_transition_list(g)
# Create data and initialize values.
node_state_grid = hmg.add_zeros('node', 'node_state_grid')
# Make the grid boundary all wall particles
node_state_grid[hmg.boundary_nodes] = 8
# Seed the grid interior with randomly oriented particles
for i in hmg.core_nodes:
if random.random() < p_init:
node_state_grid[i] = random.randint(1, 7)
# Create the CA model
ca = OrientedHexCTS(hmg, ns_dict, xn_list, node_state_grid)
# Create a CAPlotter object for handling screen display
ca_plotter = CAPlotter(ca)
# Plot the initial grid
ca_plotter.update_plot()
# RUN
current_time = 0.0
while current_time < run_duration:
# Once in a while, print out simulation and real time to let the user
# know that the sim is running ok
current_real_time = time.time()
if current_real_time >= next_report:
print('Current sim time', current_time, '(',
100 * current_time / run_duration, '%)')
next_report = current_real_time + report_interval
# Run the model forward in time until the next output step
ca.run(current_time + plot_interval,
ca.node_state,
plot_each_transition=plot_every_transition,
plotter=ca_plotter)
current_time += plot_interval
# Plot the current grid
ca_plotter.update_plot()
# FINALIZE
# Plot
ca_plotter.finalize()
def main():
# INITIALIZE
# User-defined parameters
nr = 52
nc = 120
plot_interval = 1.0
run_duration = 100.0
report_interval = 5.0 # report interval, in real-time seconds
p_init = 0.1 # probability that a cell is occupied at start
plot_every_transition = False
# Remember the clock time, and calculate when we next want to report
# progress.
current_real_time = time.time()
next_report = current_real_time + report_interval
# Create a grid
hmg = HexModelGrid(nr, nc, 1.0, orientation='vertical', reorient_links=True)
# Close the grid boundaries
#hmg.set_closed_nodes(hmg.open_boundary_nodes)
# Set up the states and pair transitions.
# Transition data here represent particles moving on a lattice: one state
# per direction (for 6 directions), plus an empty state, a stationary
# state, and a wall state.
ns_dict = { 0 : 'empty',
1 : 'moving up',
2 : 'moving right and up',
3 : 'moving right and down',
4 : 'moving down',
5 : 'moving left and down',
6 : 'moving left and up',
7 : 'rest',
8 : 'wall'}
xn_list = setup_transition_list()
# Create data and initialize values.
node_state_grid = hmg.add_zeros('node', 'node_state_grid', dtype=int)
# Make the grid boundary all wall particles
node_state_grid[hmg.boundary_nodes] = 8
# Seed the grid interior with randomly oriented particles
for i in hmg.core_nodes:
if random.random()<p_init:
node_state_grid[i] = random.randint(1, 7)
# Create the CA model
ca = OrientedHexCTS(hmg, ns_dict, xn_list, node_state_grid)
# Set up a color map for plotting
import matplotlib
clist = [ (1.0, 1.0, 1.0), # empty = white
(1.0, 0.0, 0.0), # up = red
(1.0, 1.0, 0.0), # right-up = yellow
(0.0, 1.0, 0.0), # down-up = green
(0.0, 1.0, 1.0), # down = cyan
(0.0, 0.0, 1.0), # left-down = blue
(1.0, 0.0, 1.0), # left-up = magenta
(0.5, 0.5, 0.5), # resting = gray
(0.0, 0.0, 0.0) ] # wall = black
my_cmap = matplotlib.colors.ListedColormap(clist)
# Create a CAPlotter object for handling screen display
ca_plotter = CAPlotter(ca, cmap=my_cmap)
# Plot the initial grid
ca_plotter.update_plot()
# Create an array to store the numbers of states at each plot interval
nstates = zeros((9, int(run_duration/plot_interval)))
k = 0
# RUN
current_time = 0.0
while current_time < run_duration:
# Once in a while, print out simulation and real time to let the user
# know that the sim is running ok
current_real_time = time.time()
if current_real_time >= next_report:
print('Current sim time',current_time,'(',100*current_time/run_duration,'%)')
next_report = current_real_time + report_interval
# Run the model forward in time until the next output step
ca.run(current_time+plot_interval, ca.node_state,
plot_each_transition=plot_every_transition, plotter=ca_plotter)
current_time += plot_interval
# Plot the current grid
ca_plotter.update_plot()
axis('off')
# Record numbers in each state
nstates[:,k] = bincount(node_state_grid)
k += 1
# FINALIZE
# Plot
ca_plotter.finalize()
# Display the numbers of each state
fig, ax = subplots()
for i in range(1, 8):
plot(arange(plot_interval, run_duration+plot_interval, plot_interval), nstates[i,:], label=ns_dict[i], color=clist[i])
ax.legend()
xlabel('Time')
ylabel('Number of particles in state')
title('Particle distribution by state')
axis([0, run_duration, 0, 2*nstates[7,0]])
show()
class CTSModel(object):
"""
Implement a generic CellLab-CTS model.
This is the base class from which models should inherit.
"""
def __init__(self,
grid_size=(5, 5),
report_interval=5.0,
grid_orientation="vertical",
node_layout="rect",
show_plots=False,
cts_type="oriented_hex",
run_duration=1.0,
output_interval=1.0e99,
plot_every_transition=False,
**kwds):
self.initialize(grid_size, report_interval, grid_orientation,
node_layout, show_plots, cts_type, run_duration,
output_interval, plot_every_transition, **kwds)
def initialize(self,
grid_size=(5, 5),
report_interval=5.0,
grid_orientation="vertical",
node_layout="rect",
show_plots=False,
cts_type="oriented_hex",
run_duration=1.0,
output_interval=1.0e99,
plot_every_transition=False,
**kwds):
# Remember the clock time, and calculate when we next want to report
# progress.
self.current_real_time = time.time()
self.next_report = self.current_real_time + report_interval
self.report_interval = report_interval
# Interval for output
self.output_interval = output_interval
# Duration for run
self.run_duration = run_duration
# Create a grid
self.create_grid_and_node_state_field(grid_size[0], grid_size[1],
grid_orientation, node_layout,
cts_type)
# Create the node-state dictionary
ns_dict = self.node_state_dictionary()
# Initialize values of the node-state grid
nsg = self.initialize_node_state_grid()
# Create the transition list
xn_list = self.transition_list()
# Create the CA object
if cts_type == "raster":
from landlab.ca.raster_cts import RasterCTS
self.ca = RasterCTS(self.grid, ns_dict, xn_list, nsg)
elif cts_type == "oriented_raster":
from landlab.ca.oriented_raster_cts import OrientedRasterCTS
self.ca = OrientedRasterCTS(self.grid, ns_dict, xn_list, nsg)
elif cts_type == "hex":
from landlab.ca.hex_cts import HexCTS
self.ca = HexCTS(self.grid, ns_dict, xn_list, nsg)
else:
from landlab.ca.oriented_hex_cts import OrientedHexCTS
self.ca = OrientedHexCTS(self.grid, ns_dict, xn_list, nsg)
# Initialize graphics
self._show_plots = show_plots
if show_plots:
self.initialize_plotting(**kwds)
def create_grid_and_node_state_field(self, num_rows, num_cols,
grid_orientation, node_layout,
cts_type):
"""Create the grid and the field containing node states."""
if cts_type == "raster" or cts_type == "oriented_raster":
from landlab import RasterModelGrid
self.grid = RasterModelGrid(shape=(num_rows, num_cols),
xy_spacing=1.0)
else:
from landlab import HexModelGrid
self.grid = HexModelGrid(
(num_rows, num_cols),
spacing=1.0,
orientation=grid_orientation,
node_layout=node_layout,
)
self.grid.add_zeros("node_state", at="node", dtype=int)
def node_state_dictionary(self):
"""Create and return a dictionary of all possible node (cell) states.
This method creates a default set of states (just two); it is a
template meant to be overridden.
"""
ns_dict = {0: "on", 1: "off"}
return ns_dict
def transition_list(self):
"""Create and return a list of transition objects.
This method creates a default set of transitions (just two); it is a
template meant to be overridden.
"""
xn_list = []
xn_list.append(Transition((0, 1, 0), (1, 0, 0), 1.0))
xn_list.append(Transition((1, 0, 0), (0, 1, 0), 1.0))
return xn_list
def write_output(self, grid, outfilename, iteration):
"""Write output to file (currently netCDF)."""
filename = outfilename + str(iteration).zfill(4) + ".nc"
save_grid(grid, filename)
def initialize_node_state_grid(self):
"""Initialize values in the node-state grid.
This method should be overridden. The default is random "on" and "off".
"""
num_states = 2
for i in range(self.grid.number_of_nodes):
self.grid.at_node["node_state"][i] = random.randint(num_states)
return self.grid.at_node["node_state"]
def initialize_plotting(self, **kwds):
"""Create and configure CAPlotter object."""
self.ca_plotter = CAPlotter(self.ca, **kwds)
self.ca_plotter.update_plot()
axis("off")
def run_for(self, dt):
self.ca.run(self.ca.current_time + dt, self.ca.node_state)
def main():
# INITIALIZE
# User-defined parameters
nr = 21
nc = 21
plot_interval = 0.5
run_duration = 25.0
report_interval = 5.0 # report interval, in real-time seconds
# Remember the clock time, and calculate when we next want to report
# progress.
current_real_time = time.time()
next_report = current_real_time + report_interval
# Create a grid
hmg = HexModelGrid(nr, nc, 1.0, orientation="vertical", reorient_links=True)
# Close the grid boundaries
hmg.set_closed_nodes(hmg.open_boundary_nodes)
# Set up the states and pair transitions.
# Transition data here represent the disease status of a population.
ns_dict = {0: "fluid", 1: "grain"}
xn_list = setup_transition_list()
# Create data and initialize values. We start with the 3 middle columns full
# of grains, and the others empty.
node_state_grid = hmg.add_zeros("node", "node_state_grid")
middle = 0.25 * (nc - 1) * sqrt(3)
is_middle_cols = logical_and(hmg.node_x < middle + 1., hmg.node_x > middle - 1.)
node_state_grid[where(is_middle_cols)[0]] = 1
# Create the CA model
ca = OrientedHexCTS(hmg, ns_dict, xn_list, node_state_grid)
# Create a CAPlotter object for handling screen display
ca_plotter = CAPlotter(ca)
# Plot the initial grid
ca_plotter.update_plot()
# RUN
current_time = 0.0
while current_time < run_duration:
# Once in a while, print out simulation and real time to let the user
# know that the sim is running ok
current_real_time = time.time()
if current_real_time >= next_report:
print(
"Current sim time",
current_time,
"(",
100 * current_time / run_duration,
"%)",
)
next_report = current_real_time + report_interval
# Run the model forward in time until the next output step
ca.run(current_time + plot_interval, ca.node_state, plot_each_transition=False)
current_time += plot_interval
# Plot the current grid
ca_plotter.update_plot()
# FINALIZE
# Plot
ca_plotter.finalize()
def main():
# INITIALIZE
# User-defined parameters
nr = 52
nc = 120
plot_interval = 10.0
run_duration = 1000.0
report_interval = 5.0 # report interval, in real-time seconds
p_init = 0.1 # probability that a cell is occupied at start
plot_every_transition = False
# Remember the clock time, and calculate when we next want to report
# progress.
current_real_time = time.time()
next_report = current_real_time + report_interval
# Create a grid
hmg = HexModelGrid(nr,
nc,
1.0,
orientation="vertical",
reorient_links=True)
# Set up the states and pair transitions.
# Transition data here represent particles moving on a lattice: one state
# per direction (for 6 directions), plus an empty state, a stationary
# state, and a wall state.
ns_dict = {
0: "empty",
1: "moving up",
2: "moving right and up",
3: "moving right and down",
4: "moving down",
5: "moving left and down",
6: "moving left and up",
7: "rest",
8: "wall",
}
xn_list = setup_transition_list()
# Create data and initialize values.
node_state_grid = hmg.add_zeros("node", "node_state_grid", dtype=int)
# Make the grid boundary all wall particles
node_state_grid[hmg.boundary_nodes] = 8
# Seed the grid interior with randomly oriented particles
for i in hmg.core_nodes:
if random.random() < p_init:
node_state_grid[i] = random.randint(1, 7)
# Create the CA model
ca = OrientedHexCTS(hmg, ns_dict, xn_list, node_state_grid)
# Set up a color map for plotting
import matplotlib
clist = [
(1.0, 1.0, 1.0), # empty = white
(1.0, 0.0, 0.0), # up = red
(0.8, 0.8, 0.0), # right-up = yellow
(0.0, 1.0, 0.0), # down-up = green
(0.0, 1.0, 1.0), # down = cyan
(0.0, 0.0, 1.0), # left-down = blue
(1.0, 0.0, 1.0), # left-up = magenta
(0.5, 0.5, 0.5), # resting = gray
(0.0, 0.0, 0.0),
] # wall = black
line_styles = ["", "-", "--", "--", "-", "-.", "--", "-", ":"]
my_cmap = matplotlib.colors.ListedColormap(clist)
# Create a CAPlotter object for handling screen display
ca_plotter = CAPlotter(ca, cmap=my_cmap)
# Plot the initial grid
ca_plotter.update_plot()
# Create an array to store the numbers of states at each plot interval
nstates = zeros((9, int(run_duration / plot_interval)))
k = 0
# RUN
current_time = 0.0
while current_time < run_duration:
# Once in a while, print out simulation and real time to let the user
# know that the sim is running ok
current_real_time = time.time()
if current_real_time >= next_report:
print(
"Current sim time",
current_time,
"(",
100 * current_time / run_duration,
"%)",
)
next_report = current_real_time + report_interval
# Run the model forward in time until the next output step
ca.run(
current_time + plot_interval,
ca.node_state,
plot_each_transition=plot_every_transition,
plotter=ca_plotter,
)
current_time += plot_interval
# Plot the current grid
ca_plotter.update_plot()
axis("off")
# Record numbers in each state
nstates[:, k] = bincount(node_state_grid)
k += 1
# FINALIZE
# Plot
ca_plotter.finalize()
# Display the numbers of each state
fig, ax = subplots()
for i in range(1, 8):
plot(
arange(plot_interval, run_duration + plot_interval, plot_interval),
nstates[i, :],
label=ns_dict[i],
color=clist[i],
linestyle=line_styles[i],
)
ax.legend()
xlabel("Time")
ylabel("Number of particles in state")
title("Particle distribution by state")
axis([0, run_duration, 0, 2 * nstates[7, 0]])
show()
def main():
# INITIALIZE
# User-defined parameters
nr = 80 # number of rows in grid
nc = 50 # number of columns in grid
plot_interval = 0.5 # time interval for plotting, sec
run_duration = 20.0 # duration of run, sec
report_interval = 10.0 # report interval, in real-time seconds
# Remember the clock time, and calculate when we next want to report
# progress.
current_real_time = time.time()
next_report = current_real_time + report_interval
# Create grid
mg = HexModelGrid(nr, nc, 1.0)
# Make the boundaries be walls
# mg.set_closed_boundaries_at_grid_edges(True, True, True, True)<--I am not sure what the equivalent is for hexgrid
#Create a node-state dictionary
ns_dict = {0: 'fluid', 1: 'particle'}
#Create the transition list
xn_list = setup_transition_list()
# Create the node-state array and attach it to the grid
node_state_grid = mg.add_zeros('node', 'node_state_map', dtype=int)
# Initialize the node-state array: here, the initial condition is a pile of
# resting grains at the bottom of a container.
bottom_rows = where(mg.node_y < 0.1 * nr)[0]
node_state_grid[bottom_rows] = 1
# For visual display purposes, set all boundary nodes to fluid
node_state_grid[mg.closed_boundary_nodes] = 0
# Create the CA model
ca = OrientedHexCTS(mg, ns_dict, xn_list, node_state_grid)
# Set up colors for plotting
grain = '#5F594D'
fluid = '#D0E4F2'
clist = [fluid, grain]
my_cmap = matplotlib.colors.ListedColormap(clist)
# Create a CAPlotter object for handling screen display
ca_plotter = CAPlotter(ca, cmap=my_cmap)
# Plot the initial grid
ca_plotter.update_plot()
# RUN
current_time = 0.0
while current_time < run_duration:
# Once in a while, print out simulation real time to let the user
# know that the sim is running ok
current_real_time = time.time()
if current_real_time >= next_report:
print('Current simulation time ' + str(current_time) + ' \
(' + str(int(100 * current_time / run_duration)) + '%)')
next_report = current_real_time + report_interval
# Run the model forward in time until the next output step
ca.run(current_time + plot_interval,
ca.node_state,
plot_each_transition=False)
current_time += plot_interval
# Plot the current grid
ca_plotter.update_plot()
ca_plotter.finalize()
def main():
# INITIALIZE
# User-defined parameters
nr = 80 # number of rows in grid
nc = 50 # number of columns in grid
plot_interval = 0.5 # time interval for plotting, sec
run_duration = 20.0 # duration of run, sec
report_interval = 10.0 # report interval, in real-time seconds
# Remember the clock time, and calculate when we next want to report
# progress.
current_real_time = time.time()
next_report = current_real_time + report_interval
# Create grid
mg = HexModelGrid(nr, nc, 1.0)
# Make the boundaries be walls
# mg.set_closed_boundaries_at_grid_edges(True, True, True, True)<--I am not sure what the equivalent is for hexgrid
#Create a node-state dictionary
ns_dict = { 0 : 'fluid', 1 : 'particle' }
#Create the transition list
xn_list = setup_transition_list()
# Create the node-state array and attach it to the grid
node_state_grid = mg.add_zeros('node', 'node_state_map', dtype=int)
# Initialize the node-state array: here, the initial condition is a pile of
# resting grains at the bottom of a container.
bottom_rows = where(mg.node_y<0.1*nr)[0]
node_state_grid[bottom_rows] = 1
# For visual display purposes, set all boundary nodes to fluid
node_state_grid[mg.closed_boundary_nodes] = 0
# Create the CA model
ca = OrientedHexCTS(mg, ns_dict, xn_list, node_state_grid)
# Set up colors for plotting
grain = '#5F594D'
fluid = '#D0E4F2'
clist = [fluid,grain]
my_cmap = matplotlib.colors.ListedColormap(clist)
# Create a CAPlotter object for handling screen display
ca_plotter = CAPlotter(ca, cmap=my_cmap)
# Plot the initial grid
ca_plotter.update_plot()
# RUN
current_time = 0.0
while current_time < run_duration:
# Once in a while, print out simulation real time to let the user
# know that the sim is running ok
current_real_time = time.time()
if current_real_time >= next_report:
print('Current simulation time '+str(current_time)+' \
('+str(int(100*current_time/run_duration))+'%)')
next_report = current_real_time + report_interval
# Run the model forward in time until the next output step
ca.run(current_time+plot_interval, ca.node_state, plot_each_transition=False)
current_time += plot_interval
# Plot the current grid
ca_plotter.update_plot()
ca_plotter.finalize()
class CTSModel(object):
"""
Implement a generic CellLab-CTS model.
This is the base class from which models should inherit.
"""
def __init__(self,
grid_size=(5, 5),
report_interval=5.0,
grid_orientation='vertical',
node_layout='rect',
show_plots=False,
cts_type='oriented_hex',
run_duration=1.0,
output_interval=1.0e99,
plot_every_transition=False,
initial_state_grid=None,
prop_data=None,
prop_reset_value=None,
seed=0,
closed_boundaries=(False, False, False, False),
**kwds):
self.initialize(grid_size, report_interval, grid_orientation,
node_layout, show_plots, cts_type, run_duration,
output_interval, plot_every_transition,
initial_state_grid, prop_data, prop_reset_value, seed,
closed_boundaries, **kwds)
def initialize(self,
grid_size=(5, 5),
report_interval=5.0,
grid_orientation='vertical',
node_layout='rect',
show_plots=False,
cts_type='oriented_hex',
run_duration=1.0,
output_interval=1.0e99,
plot_every_transition=False,
initial_state_grid=None,
prop_data=None,
prop_reset_value=None,
seed=0,
closed_boundaries=(False, False, False, False),
**kwds):
"""Initialize CTSModel."""
# Remember the clock time, and calculate when we next want to report
# progress.
self.current_real_time = time.time()
self.next_report = self.current_real_time + report_interval
self.report_interval = report_interval
# Interval for output
self.output_interval = output_interval
# Duration for run
self.run_duration = run_duration
# Create a grid
self.create_grid_and_node_state_field(grid_size[0], grid_size[1],
grid_orientation, node_layout,
cts_type, closed_boundaries)
# If prop_data is a string, we assume it is a field name
if isinstance(prop_data, string_types):
prop_data = self.grid.add_zeros('node', prop_data)
# Create the node-state dictionary
ns_dict = self.node_state_dictionary()
# Initialize values of the node-state grid
if initial_state_grid is None:
nsg = self.initialize_node_state_grid()
else:
try:
nsg = initial_state_grid
self.grid.at_node['node_state'][:] = nsg
except TypeError:
print('If initial_state_grid given, must be array of int')
raise
# Create the transition list
xn_list = self.transition_list()
# Create the CA object
if cts_type == 'raster':
from landlab.ca.raster_cts import RasterCTS
self.ca = RasterCTS(self.grid,
ns_dict,
xn_list,
nsg,
prop_data,
prop_reset_value,
seed=seed)
elif cts_type == 'oriented_raster':
from landlab.ca.oriented_raster_cts import OrientedRasterCTS
self.ca = OrientedRasterCTS(self.grid,
ns_dict,
xn_list,
nsg,
prop_data,
prop_reset_value,
seed=seed)
elif cts_type == 'hex':
from landlab.ca.hex_cts import HexCTS
self.ca = HexCTS(self.grid,
ns_dict,
xn_list,
nsg,
prop_data,
prop_reset_value,
seed=seed)
else:
from landlab.ca.oriented_hex_cts import OrientedHexCTS
self.ca = OrientedHexCTS(self.grid,
ns_dict,
xn_list,
nsg,
prop_data,
prop_reset_value,
seed=seed)
# Initialize graphics
self._show_plots = show_plots
if show_plots:
self.initialize_plotting(**kwds)
def _set_closed_boundaries_for_hex_grid(self, closed_boundaries):
"""Setup one or more closed boundaries for a hex grid.
Parameters
----------
closed_boundaries : 4-element tuple of bool\
Whether right, top, left, and bottom edges have closed nodes
Examples
--------
>>> from grainhill import CTSModel
>>> cm = CTSModel(closed_boundaries=(True, True, True, True))
>>> cm.grid.status_at_node # doctest: +NORMALIZE_WHITESPACE
array([4, 4, 4, 4, 4, 4, 0, 4, 0, 0, 4, 0, 4, 0, 0, 4, 0, 4, 0, 0, 4,
4, 4, 4, 4], dtype=uint8)
"""
g = self.grid
if closed_boundaries[0]:
g.status_at_node[g.nodes_at_right_edge] = g.BC_NODE_IS_CLOSED
if closed_boundaries[1]:
g.status_at_node[g.nodes_at_top_edge] = g.BC_NODE_IS_CLOSED
if closed_boundaries[2]:
g.status_at_node[g.nodes_at_left_edge] = g.BC_NODE_IS_CLOSED
if closed_boundaries[3]:
g.status_at_node[g.nodes_at_bottom_edge] = g.BC_NODE_IS_CLOSED
def create_grid_and_node_state_field(self, num_rows, num_cols,
grid_orientation, node_layout,
cts_type, closed_bounds):
"""Create the grid and the field containing node states."""
if cts_type == 'raster' or cts_type == 'oriented_raster':
from landlab import RasterModelGrid
self.grid = RasterModelGrid(shape=(num_rows, num_cols),
spacing=1.0)
self.grid.set_closed_boundaries_at_grid_edges(
closed_bounds[0], closed_bounds[1], closed_bounds[2],
closed_bounds[3])
else:
from landlab import HexModelGrid
self.grid = HexModelGrid(shape=(num_rows, num_cols),
spacing=1.0,
orientation=grid_orientation,
node_layout=node_layout)
if True in closed_bounds:
self._set_closed_boundaries_for_hex_grid(closed_bounds)
self.grid.add_zeros('node', 'node_state', dtype=int)
def node_state_dictionary(self):
"""Create and return a dictionary of all possible node (cell) states.
This method creates a default set of states (just two); it is a
template meant to be overridden.
"""
ns_dict = {0: 'on', 1: 'off'}
return ns_dict
def transition_list(self):
"""Create and return a list of transition objects.
This method creates a default set of transitions (just two); it is a
template meant to be overridden.
"""
xn_list = []
xn_list.append(Transition((0, 1, 0), (1, 0, 0), 1.0))
xn_list.append(Transition((1, 0, 0), (0, 1, 0), 1.0))
return xn_list
def write_output(self, grid, outfilename, iteration):
"""Write output to file (currently netCDF)."""
filename = outfilename + str(iteration).zfill(4) + '.nc'
save_grid(grid, filename)
def initialize_node_state_grid(self):
"""Initialize values in the node-state grid.
This method should be overridden. The default is random "on" and "off".
"""
num_states = 2
for i in range(self.grid.number_of_nodes):
self.grid.at_node['node_state'][i] = random.randint(num_states)
return self.grid.at_node['node_state']
def initialize_plotting(self, **kwds):
"""Create and configure CAPlotter object."""
self.ca_plotter = CAPlotter(self.ca, **kwds)
self.ca_plotter.update_plot()
axis('off')
def run_for(self, dt):
self.ca.run(self.ca.current_time + dt, self.ca.node_state)
