def test_output_filtering(self):
"""
Tests whether the 'outputs' argument is able to filter (presumably)
unconnected nonlinear mappings from inputs and states to analog send
ports
"""
e = Dynamics(name='E',
regimes=[
Regime('dSV1/dt = -SV1 / P1',
'dSV2/dt = -SV2 / P1 + ARP1 * P2',
name='R1')
],
aliases=['A1:=SV1 * C1', 'A2:=SV1 * SV2'],
analog_ports=[
AnalogReceivePort('ARP1', dimension=un.per_time),
AnalogSendPort('A1', dimension=un.current),
AnalogSendPort('A2', dimension=un.dimensionless)
],
parameters=[
Parameter('P1', dimension=un.time),
Parameter('P2', dimension=un.dimensionless)
],
constants=[Constant('C1', 10, units=un.mA)])
self.assertFalse(e.is_linear())
self.assertTrue(e.is_linear(outputs=['A1']))
def test_nonlinear_state_assignment_in_onevent(self):
"""Test that nonlinear state assignements in on events"""
g = Dynamics(name='G',
regimes=[
Regime('dSV1/dt = -SV1 / P1',
'dSV2/dt = -SV2 / P1 + ARP1 * P2',
name='R1',
transitions=[
OnEvent('ERP1',
state_assignments=[
StateAssignment(
'SV2', 'SV2 + A2')
])
])
],
aliases=['A1:=SV1 * C1', 'A2:=P3 * P4 / ADP1'],
analog_ports=[
AnalogReceivePort('ARP1', dimension=un.per_time),
AnalogReducePort('ADP1',
operator='+',
dimension=un.dimensionless),
AnalogSendPort('A1', dimension=un.current),
AnalogSendPort('A2', dimension=un.dimensionless)
],
event_ports=[EventReceivePort('ERP1')],
parameters=[
Parameter('P1', dimension=un.time),
Parameter('P2', dimension=un.dimensionless),
Parameter('P3', dimension=un.resistance),
Parameter('P4', dimension=un.conductance)
],
constants=[Constant('C1', 10, units=un.nA)])
self.assertFalse(g.is_linear())
def test_output_filtering(self):
"""
Tests whether the 'outputs' argument is able to filter (presumably)
unconnected nonlinear mappings from inputs and states to analog send
ports
"""
e = Dynamics(
name='E',
regimes=[
Regime('dSV1/dt = -SV1 / P1',
'dSV2/dt = -SV2 / P1 + ARP1 * P2', name='R1')],
aliases=['A1:=SV1 * C1', 'A2:=SV1 * SV2'],
analog_ports=[AnalogReceivePort('ARP1', dimension=un.per_time),
AnalogSendPort('A1', dimension=un.current),
AnalogSendPort('A2', dimension=un.dimensionless)],
parameters=[Parameter('P1', dimension=un.time),
Parameter('P2', dimension=un.dimensionless)],
constants=[Constant('C1', 10, units=un.mA)])
self.assertFalse(e.is_linear())
self.assertTrue(e.is_linear(outputs=['A1']))
def test_input_multiplication_nonlinear(self):
"""Nonlinear due to multiplication of SV1 and SV2 in SV1 T.D."""
d = Dynamics(
name='D',
regimes=[
Regime('dSV1/dt = -SV1 * SV2 / P1',
'dSV2/dt = -SV2 / P1 + ARP1 * P2', name='R1')],
analog_ports=[AnalogReceivePort('ARP1', dimension=un.per_time)],
parameters=[Parameter('P1', dimension=un.time),
Parameter('P2', dimension=un.dimensionless)])
self.assertFalse(d.is_linear())
def test_nonlinear_function(self):
"""Nonlinear due function in SV1 T.D."""
h = Dynamics(
name='H',
regimes=[
Regime('dSV1/dt = -sin(SV1) / P1',
'dSV2/dt = -SV2 / P1 + ARP1 * P2', name='R1')],
analog_ports=[AnalogReceivePort('ARP1', dimension=un.per_time)],
parameters=[Parameter('P1', dimension=un.time),
Parameter('P2', dimension=un.dimensionless)])
self.assertFalse(h.is_linear())
def test_basic_linear(self):
"""A basic linear dynamics example"""
a = Dynamics(
name='A',
regimes=[
Regime('dSV1/dt = -SV1 / P1',
'dSV2/dt = -SV2 / P1 + ARP1 * P2', name='R1')],
analog_ports=[AnalogReceivePort('ARP1', dimension=un.per_time)],
parameters=[Parameter('P1', dimension=un.time),
Parameter('P2', dimension=un.dimensionless)])
self.assertTrue(a.is_linear())
def test_multi_regime_nonlinear(self):
"""Nonlinear due to multiple regimes"""
b = Dynamics(
name='B',
regimes=[
Regime('dSV1/dt = -SV1 / P1',
transitions=[OnEvent('ERP1', target_regime_name='R2')],
name='R1'),
Regime('dSV1/dt = -SV1 / P1',
transitions=[OnEvent('ERP1', target_regime_name='R1')],
name='R2')],
event_ports=[EventReceivePort('ERP1')],
parameters=[Parameter('P1', dimension=un.time)])
self.assertFalse(b.is_linear())
def test_on_condition_state_assignment_nonlinear(self):
"""
Nonlinear due to a state assignment in on-condition (i.e. piece-wise
dynamics
"""
c = Dynamics(
name='C',
regimes=[
Regime('dSV1/dt = -P1 * SV1',
transitions=[OnCondition(
'SV1 > 10',
state_assignments=[StateAssignment('SV1', 20)])],
name='R1')],
parameters=[Parameter('P1', dimension=un.per_time)])
self.assertFalse(c.is_linear())
def test_basic_linear(self):
"""A basic linear dynamics example"""
a = Dynamics(
name='A',
regimes=[
Regime('dSV1/dt = -SV1 / P1',
'dSV2/dt = -SV2 / P1 + ARP1 * P2',
name='R1')
],
analog_ports=[AnalogReceivePort('ARP1', dimension=un.per_time)],
parameters=[
Parameter('P1', dimension=un.time),
Parameter('P2', dimension=un.dimensionless)
])
self.assertTrue(a.is_linear())
def test_input_multiplication_nonlinear(self):
"""Nonlinear due to multiplication of SV1 and SV2 in SV1 T.D."""
d = Dynamics(
name='D',
regimes=[
Regime('dSV1/dt = -SV1 * SV2 / P1',
'dSV2/dt = -SV2 / P1 + ARP1 * P2',
name='R1')
],
analog_ports=[AnalogReceivePort('ARP1', dimension=un.per_time)],
parameters=[
Parameter('P1', dimension=un.time),
Parameter('P2', dimension=un.dimensionless)
])
self.assertFalse(d.is_linear())
def test_multi_regime_nonlinear(self):
"""Nonlinear due to multiple regimes"""
b = Dynamics(
name='B',
regimes=[
Regime('dSV1/dt = -SV1 / P1',
transitions=[OnEvent('ERP1', target_regime_name='R2')],
name='R1'),
Regime('dSV1/dt = -SV1 / P1',
transitions=[OnEvent('ERP1', target_regime_name='R1')],
name='R2')
],
event_ports=[EventReceivePort('ERP1')],
parameters=[Parameter('P1', dimension=un.time)])
self.assertFalse(b.is_linear())
def test_nonlinear_function(self):
"""Nonlinear due function in SV1 T.D."""
h = Dynamics(
name='H',
regimes=[
Regime('dSV1/dt = -sin(SV1) / P1',
'dSV2/dt = -SV2 / P1 + ARP1 * P2',
name='R1')
],
analog_ports=[AnalogReceivePort('ARP1', dimension=un.per_time)],
parameters=[
Parameter('P1', dimension=un.time),
Parameter('P2', dimension=un.dimensionless)
])
self.assertFalse(h.is_linear())
def test_on_condition_state_assignment_nonlinear(self):
"""
Nonlinear due to a state assignment in on-condition (i.e. piece-wise
dynamics
"""
c = Dynamics(name='C',
regimes=[
Regime('dSV1/dt = -P1 * SV1',
transitions=[
OnCondition('SV1 > 10',
state_assignments=[
StateAssignment('SV1', 20)
])
],
name='R1')
],
parameters=[Parameter('P1', dimension=un.per_time)])
self.assertFalse(c.is_linear())
def test_linear_state_assignment_in_onevent(self):
"""Test linear state assignments in on events"""
f = Dynamics(
name='F',
regimes=[
Regime('dSV1/dt = -SV1 / P1',
'dSV2/dt = -SV2 / P1 + ARP1 * P2', name='R1',
transitions=[
OnEvent('ERP1', state_assignments=[
StateAssignment('SV2', 'SV2 + A2')])])],
aliases=['A1:=SV1 * C1', 'A2:=P3 * P4'],
analog_ports=[AnalogReceivePort('ARP1', dimension=un.per_time),
AnalogSendPort('A1', dimension=un.current),
AnalogSendPort('A2', dimension=un.dimensionless)],
event_ports=[EventReceivePort('ERP1')],
parameters=[Parameter('P1', dimension=un.time),
Parameter('P2', dimension=un.dimensionless),
Parameter('P3', dimension=un.resistance),
Parameter('P4', dimension=un.conductance)],
constants=[Constant('C1', 10, units=un.pA)])
self.assertTrue(f.is_linear())
