88cd16a4e3c70d82567c96c1ffc277ef8ea87a3f,pymc3/tests/test_hmc.py,,test_leapfrog_reversible,#,14

Before Change

            q, p = q0, p0
            q, p, _ = step.leapfrog(q, p, floatX(np.array(epsilon)), np.array(n_steps, dtype="int32"))
            q, p, _ = step.leapfrog(q, -p, floatX(np.array(epsilon)), np.array(n_steps, dtype="int32"))
            close_to(q, q0, precision, str((n_steps, epsilon)))
            close_to(-p, p0, precision, str((n_steps, epsilon)))

After Change

    np.random.seed(42)
    start, model, _ = models.non_normal(n)
    size = model.ndim
    scaling = floatX(np.random.randn(size) ** 2)
    step = BaseHMC(vars=model.vars, model=model, scaling=scaling)
    step.integrator._logp_dlogp_func.set_extra_values({})
    p = floatX(step.potential.random())
    q = floatX(np.random.randn(size))
    start = step.integrator.compute_state(p, q)
    precision = select_by_precision(float64=1E-8, float32=1E-4)
    for epsilon in [.01, .1, 1.2]:
        for n_steps in [1, 2, 3, 4, 20]:
            state = start
            for _ in range(n_steps):
                state = step.integrator.step(epsilon, state)
            for _ in range(n_steps):
                state = step.integrator.step(-epsilon, state)
            close_to(state.q, start.q, precision, str((n_steps, epsilon)))
            close_to(state.p, start.p, precision, str((n_steps, epsilon)))

Instances