def test_plot_coeffs():
with pytest.raises(Exception) as excinfo:
Optimize.plot_coeffs([-0., 5., 5., -0., 5., 5.], [1, 2], [10, 15, 20])
assert "do not match" in str(excinfo.value)
#assert(type(Optimize.plot_coeffs([-0., 5., 5., -0., 5., 5.], [1,2,3], [10,15,20])) == plt)
assert (Optimize.plot_coeffs([-0., 5., 5., -0., 5., 5.], [1, 2, 3],
[10, 15, 20]) == True)
def test_GradDescent():
assert (np.round(
Optimize.grad_descent('x^2', 1.0, 0.0001, max_iters=5000, eta=0.5)[0],
3) == 0.00)
assert (np.round(
Optimize.grad_descent('x^2 - 1', 1, 0.0001, max_iters=5000,
eta=0.5)[0], 3) == round(0.000, 3))
assert (np.round(
Optimize.grad_descent('x^2 - 2', 2.0, 0.001, max_iters=5000,
eta=0.5)[0], 3) == round(0.000, 3))
def test_Newton():
assert (np.round(
Optimize.newtons_method('x^2 - 1', 0.25, 0.0001, max_iters=5000)[0],
3) == 1.000)
assert (np.round(
Optimize.newtons_method('x^2 - 2', 2.0, 0.001, max_iters=500)[0],
3) == round(math.sqrt(2), 3))
assert (np.round(
Optimize.newtons_method('x^3 - 3*x + 1', 2.0, 0.001, max_iters=500)[0],
3) == 1.532)
def test_quadratic_spline():
assert (int(abs(Optimize.quadratic_spline([2, 3, 4], [3, 8, 11])[0])) == 0)
assert (int(
Optimize.quadratic_spline([2, 3, 4, 5, 6],
[3, 8, 11, 14, 18])[1]) == 5)
with pytest.raises(Exception) as excinfo:
Optimize.quadratic_spline([1, 2], [10, 15, 20])
assert "do not match" in str(excinfo.value)
values = Optimize.quadratic_spline([2, 3, 4, 5, 6],
[-3, -8, -11, -14, -18])
assert (int(values[0] * 4 + values[1] * 3 + values[2]) == -8)
def test_animate_newtons():
f = 'x^3 - 3*x^2 + 4'
assert (Optimize.animate_newtons(f,
0.3,
epsilon=0.000001,
max_iters=500,
runtime=20) == None)
def test_animate_gradient():
f = 'tan(sin(x) + 3)'
assert (Optimize.animate_grad_desc(f,
4,
epsilon=0.00001,
max_iters=500,
eta=0.1,
runtime=20,
method='grad') == None)