def transform_shapes_and_odes(self, neuron, shape_to_buffers):
# type: (ASTNeuron, map(str, str)) -> ASTNeuron
"""
Solves all odes and equations in the handed over neuron.
Precondition: it should be ensured that most one equations block is present.
:param neuron: a single neuron instance.
:param shape_to_buffers: Map of shape names to buffers to which they were connected.
:return: A transformed version of the neuron that can be passed to the GSL.
"""
assert isinstance(
neuron.get_equations_blocks(), ASTEquationsBlock
), "Precondition violated: only one equation block should be present"
equations_block = neuron.get_equations_block()
if len(equations_block.get_ode_shapes()) == 0:
code, message = Messages.get_neuron_solved_by_solver(
neuron.get_name())
Logger.log_message(neuron, code, message,
neuron.get_source_position(), LoggingLevel.INFO)
return neuron
elif len(equations_block.get_ode_shapes()) == 1 and \
str(equations_block.get_ode_shapes()[0].get_expression()).strip().startswith(
"delta"): # assume the model is well formed
shape = equations_block.get_ode_shapes()[0]
integrate_delta_solution(equations_block, neuron, shape,
shape_to_buffers)
return neuron
elif len(equations_block.get_ode_equations()) == 1:
code, message = Messages.get_neuron_analyzed(neuron.get_name())
Logger.log_message(neuron, code, message,
neuron.get_source_position(), LoggingLevel.INFO)
solver_result = self.solve_ode_with_shapes(equations_block)
if solver_result["solver"] is "analytical":
neuron = integrate_exact_solution(neuron, solver_result)
neuron.remove_equations_block()
elif (solver_result["solver"] is "numeric"
and self.is_functional_shape_present(
equations_block.get_ode_shapes())):
functional_shapes_to_odes(neuron, solver_result)
return neuron
else:
code, message = Messages.get_neuron_solved_by_solver(
neuron.get_name())
Logger.log_message(neuron, code, message,
neuron.get_source_position(), LoggingLevel.INFO)
if self.is_functional_shape_present(
equations_block.get_ode_shapes()):
ode_shapes = self.solve_functional_shapes(equations_block)
functional_shapes_to_odes(neuron, ode_shapes)
return neuron
def transform_shapes_and_odes(neuron, shape_to_buffers):
# type: (ASTNeuron, map(str, str)) -> ASTNeuron
"""
Solves all odes and equations in the handed over neuron.
:param neuron: a single neuron instance.
:param shape_to_buffers: Map of shape names to buffers to which they were connected.
:return: A transformed version of the neuron that can be passed to the GSL.
"""
# it should be ensured that most one equations block is present
result = neuron
if isinstance(neuron.get_equations_blocks(), ASTEquationsBlock):
equations_block = neuron.get_equations_block()
if len(equations_block.get_ode_shapes()) == 0:
code, message = Messages.get_neuron_solved_by_solver(neuron.get_name())
Logger.log_message(neuron, code, message, neuron.get_source_position(), LoggingLevel.INFO)
result = neuron
if len(equations_block.get_ode_shapes()) == 1 and \
str(equations_block.get_ode_shapes()[0].get_expression()).strip().startswith(
"delta"): # assume the model is well formed
shape = equations_block.get_ode_shapes()[0]
integrate_delta_solution(equations_block, neuron, shape, shape_to_buffers)
return result
elif len(equations_block.get_ode_equations()) == 1:
code, message = Messages.get_neuron_analyzed(neuron.get_name())
Logger.log_message(neuron, code, message, neuron.get_source_position(), LoggingLevel.INFO)
solver_result = solve_ode_with_shapes(equations_block)
if solver_result["solver"] is "analytical":
result = integrate_exact_solution(neuron, solver_result)
result.remove_equations_block()
elif solver_result["solver"] is "numeric":
at_least_one_functional_shape = False
for shape in equations_block.get_ode_shapes():
if shape.get_variable().get_differential_order() == 0:
at_least_one_functional_shape = True
if at_least_one_functional_shape:
functional_shapes_to_odes(result, solver_result)
else:
result = neuron
else:
code, message = Messages.get_neuron_solved_by_solver(neuron.get_name())
Logger.log_message(neuron, code, message, neuron.get_source_position(), LoggingLevel.INFO)
at_least_one_functional_shape = False
for shape in equations_block.get_ode_shapes():
if shape.get_variable().get_differential_order() == 0:
at_least_one_functional_shape = True
break
if at_least_one_functional_shape:
ode_shapes = solve_functional_shapes(equations_block)
functional_shapes_to_odes(result, ode_shapes)
apply_spikes_from_buffers(result, shape_to_buffers)
return result
def transform_shapes_and_odes(self, neuron, shape_to_buffers):
# type: (ASTNeuron, map(str, str)) -> ASTNeuron
"""
Solves all odes and equations in the handed over neuron.
Precondition: it should be ensured that most one equations block is present.
:param neuron: a single neuron instance.
:param shape_to_buffers: Map of shape names to buffers to which they were connected.
:return: A transformed version of the neuron that can be passed to the GSL.
"""
assert isinstance(neuron.get_equations_blocks(), ASTEquationsBlock), "Precondition violated: only one equation block should be present"
equations_block = neuron.get_equations_block()
if len(equations_block.get_ode_shapes()) == 0:
code, message = Messages.get_neuron_solved_by_solver(neuron.get_name())
Logger.log_message(neuron, code, message, neuron.get_source_position(), LoggingLevel.INFO)
return neuron
elif len(equations_block.get_ode_shapes()) == 1 and \
str(equations_block.get_ode_shapes()[0].get_expression()).strip().startswith(
"delta"): # assume the model is well formed
shape = equations_block.get_ode_shapes()[0]
integrate_delta_solution(equations_block, neuron, shape, shape_to_buffers)
return neuron
elif len(equations_block.get_ode_equations()) == 1:
code, message = Messages.get_neuron_analyzed(neuron.get_name())
Logger.log_message(neuron, code, message, neuron.get_source_position(), LoggingLevel.INFO)
solver_result = self.solve_ode_with_shapes(equations_block)
if solver_result["solver"] is "analytical":
neuron = integrate_exact_solution(neuron, solver_result)
neuron.remove_equations_block()
elif (solver_result["solver"] is "numeric"
and self.is_functional_shape_present(equations_block.get_ode_shapes())):
functional_shapes_to_odes(neuron, solver_result)
return neuron
else:
code, message = Messages.get_neuron_solved_by_solver(neuron.get_name())
Logger.log_message(neuron, code, message, neuron.get_source_position(), LoggingLevel.INFO)
if self.is_functional_shape_present(equations_block.get_ode_shapes()):
ode_shapes = self.solve_functional_shapes(equations_block)
functional_shapes_to_odes(neuron, ode_shapes)
return neuron
