Exemple #1
0
def test_c_se_sim():

    c = new("C", value=1)
    se = new("Se")
    r = new("R", value=1)
    kcl = new("1")
    bg = new()
    bg.add([c, se, kcl, r])

    connect(c, (kcl, kcl.non_inverting))
    connect(r, (kcl, kcl.non_inverting))
    connect(se, (kcl, kcl.non_inverting))

    def u(t, x, dx):
        return -np.exp(-t)
    assert str(bg.constitutive_relations) == '[dx_0 + u_0 + x_0]'
    with pytest.raises(ModelException) as ex:
        _ = simulate(
            c, timespan=[0, 10], x0=[0]
        )
        assert "Control variable u_0 must be specified" in ex.args
    t, x = simulate(
        bg, timespan=[0, 10], x0=[0], control_vars=[u]
    )

    assert t[0] == 0
    assert t[-1] == 10

    assert (len(t), 1) == x.shape
    assert x[0, 0] == 0
    solution = t * np.exp(-t)

    res = abs(x - solution)
    assert (res < 0.001).all()
Exemple #2
0
def test_c_se_sim():

    c = new("C", value=1)
    se = new("Se")
    r = new("R", value=1)
    kcl = new("1")
    bg = new()
    bg.add([c, se, kcl, r])

    connect(c, (kcl, kcl.non_inverting))
    connect(r, (kcl, kcl.non_inverting))
    connect(se, (kcl, kcl.non_inverting))

    with pytest.raises(ModelException) as ex:
        t, x = simulate(c, timespan=[0, 10], x0=[0], dx0=[1])
        assert "Control variable not specified" in ex.args
    t, x = simulate(bg,
                    timespan=[0, 10],
                    x0=[0],
                    dx0=[1],
                    control_vars=['-exp(-t)'])

    assert t[0] == 0
    assert t[-1] == 10
    t_cols, _ = t.shape
    assert t_cols > 1
    assert (t_cols, 1) == x.shape
    assert x[0, 0] == 0

    solution = (1 + t) * np.exp(-t)
    solution = solution.reshape(t_cols, 1)

    assert ((x - solution) < 0.001).all()
Exemple #3
0
def test_rlc():
    c = new("C", value=1)
    se = new("Se")
    r = new("R", value=1)
    l = new("I", value=1)
    kvl = new("0")

    bg = new()
    bg.add([c, se, kvl, r, l])

    connect(c, kvl)
    connect(r, kvl)
    connect(se, kvl)
    connect(l, kvl)

    t, x = simulate(bg, timespan=[0, 10], x0=[1, 0], control_vars=[1])
Exemple #4
0
def test_closed_cycle():
    model = biochemical_cycle()

    x0 = [2.0, 2.0, 2.0]
    tspan = (0, 1.0)
    K_X_vals = [1.0, 2.0, 3.0, 4.0]

    r1 = (model / "R:r1").params['r']['value']
    K_Y = (model / "C:Y").params['k']['value']
    def v1(r, K_X, K_Y, x_X, x_Y): return r * K_X * x_X - r * K_Y * x_Y

    for K_X in K_X_vals:
        (model / "C:X").set_param('k', K_X)
        t, x = simulate(model, tspan, x0, dt=0.01)
        # Check initial reaction rate
        assert v1(r1, K_X, K_Y, x[0][0], x[0][1]) == -4 + 2 * K_X
        # Check that the final reaction rate is small
        assert v1(r1, K_X, K_Y, x[-1][0], x[-1][1]) < 1e-4
Exemple #5
0
def test_c_se_sum_switch():
    c = new("C", value=1)
    se = new("Se")
    r = new("R", value=1)
    kcl = new("1")
    bg = new()
    bg.add([c, se, kcl, r])

    connect(c, (kcl, kcl.non_inverting))
    connect(r, (kcl, kcl.non_inverting))
    connect(se, (kcl, kcl.non_inverting))

    def bang_bang(t, x, dx):
        return 1.5 if x >= 1 else -2.0

    t, x = simulate(
        bg, timespan=[0, 10], x0=[0], dx0=[1], control_vars=[bang_bang])

    assert (x[0, -1] - 1) < 0.001
Exemple #6
0
def test_c_se_sum_switch():
    c = new("C", value=1)
    se = new("Se")
    r = new("R", value=1)
    kcl = new("1")
    bg = new()
    bg.add([c, se, kcl, r])

    connect(c, (kcl, kcl.non_inverting))
    connect(r, (kcl, kcl.non_inverting))
    connect(se, (kcl, kcl.non_inverting))

    bang_bang = ["x_0 >= 1 ?  1.5: -2 "]

    t, x = simulate(bg,
                    timespan=[0, 10],
                    x0=[0],
                    dx0=[1],
                    control_vars=bang_bang)

    assert (x[0, -1] - 1) < 0.001
Exemple #7
0
def test_open_cycle():
    model = open_cycle()

    x0 = [2.0,2.0,2.0]
    tspan = (0,10.0)
    K_A = 1
    x_A_vals = [0.02,2.0,200]
    expected_fluxes = [-3.0906200316489003,0.0,10.860335195530652]

    r1 = (model/"R:r1").params['r']['value']
    K_X = (model/"C:X").params['k']['value']
    K_Y = (model/"C:Y").params['k']['value']
    def v1(r,K_X,K_Y,K_A,x_X,x_Y,x_A): return r*K_X*x_X*K_A*x_A - r*K_Y*x_Y

    for x_A,v_true in zip(x_A_vals,expected_fluxes):
        mu_A = np.log(K_A*x_A)
        (model/"SS:A").set_param('e',mu_A)
        t,x = simulate(model,tspan,x0)
        
        x_end = x[-1]
        flux = v1(r1,K_X,K_Y,K_A,x_end[0],x_end[1],x_A)
        assert flux == pytest.approx(v_true,abs=1e-6)
Exemple #8
0
def test_c_sim_fail():

    c = new("C")
    with pytest.raises(ModelException):
        t, x = simulate(c, timespan=[0, 1], x0=[1], dx0=[1])