Replace uses of testval with initval

Author: brandonwillard

Diff for: pymc3/distributions/bart.py

@@ -27,7 +27,7 @@ def __init__(self, X, Y, m=200, alpha=0.25, split_prior=None, *args, **kwargs):
 
         self.X, self.Y, self.missing_data = self.preprocess_XY(X, Y)
 
-        super().__init__(shape=X.shape[0], dtype="float64", testval=0, *args, **kwargs)
+        super().__init__(shape=X.shape[0], dtype="float64", initval=0, *args, **kwargs)
 
         if self.X.ndim != 2:
             raise ValueError("The design matrix X must have two dimensions")

Diff for: pymc3/distributions/bound.py

@@ -42,7 +42,7 @@ def __init__(self, distribution, lower, upper, default, *args, **kwargs):
         super().__init__(
             shape=self._wrapped.shape,
             dtype=self._wrapped.dtype,
-            testval=self._wrapped.testval,
+            initval=self._wrapped.initval,
             defaults=defaults,
             transform=self._wrapped.transform,
         )
@@ -252,15 +252,15 @@ class Bound:
 
         with pm.Model():
             NegativeNormal = pm.Bound(pm.Normal, upper=0.0)
-            par1 = NegativeNormal('par`', mu=0.0, sigma=1.0, testval=-0.5)
+            par1 = NegativeNormal('par`', mu=0.0, sigma=1.0, initval=-0.5)
             # you can use the Bound object multiple times to
             # create multiple bounded random variables
-            par1_1 = NegativeNormal('par1_1', mu=-1.0, sigma=1.0, testval=-1.5)
+            par1_1 = NegativeNormal('par1_1', mu=-1.0, sigma=1.0, initval=-1.5)
 
             # you can also define a Bound implicitly, while applying
             # it to a random variable
             par2 = pm.Bound(pm.Normal, lower=-1.0, upper=1.0)(
-                    'par2', mu=0.0, sigma=1.0, testval=1.0)
+                    'par2', mu=0.0, sigma=1.0, initval=1.0)
     """
 
     def __init__(self, distribution, lower=None, upper=None):

Diff for: pymc3/distributions/distribution.py

@@ -277,14 +277,14 @@ def __init__(
         self,
         shape,
         dtype,
-        testval=None,
+        initval=None,
         defaults=(),
         parent_dist=None,
         *args,
         **kwargs,
     ):
         super().__init__(
-            shape=shape, dtype=dtype, testval=testval, defaults=defaults, *args, **kwargs
+            shape=shape, dtype=dtype, initval=initval, defaults=defaults, *args, **kwargs
         )
         self.parent_dist = parent_dist
 
@@ -342,7 +342,7 @@ def __init__(
         logp,
         shape=(),
         dtype=None,
-        testval=0,
+        initval=0,
         random=None,
         wrap_random_with_dist_shape=True,
         check_shape_in_random=True,
@@ -363,8 +363,8 @@ def __init__(
             a value here.
         dtype: None, str (Optional)
             The dtype of the distribution.
-        testval: number or array (Optional)
-            The ``testval`` of the RV's tensor that follow the ``DensityDist``
+        initval: number or array (Optional)
+            The ``initval`` of the RV's tensor that follow the ``DensityDist``
             distribution.
         args, kwargs: (Optional)
             These are passed to the parent class' ``__init__``.
@@ -400,7 +400,7 @@ def __init__(
         """
         if dtype is None:
             dtype = aesara.config.floatX
-        super().__init__(shape, dtype, testval, *args, **kwargs)
+        super().__init__(shape, dtype, initval, *args, **kwargs)
         self.logp = logp
         if type(self.logp) == types.MethodType:
             if PLATFORM != "linux":

Diff for: pymc3/distributions/mixture.py

@@ -609,7 +609,7 @@ class NormalMixture(Mixture):
                 10,
                 shape=n_components,
                 transform=pm.transforms.ordered,
-                testval=[1, 2, 3],
+                initval=[1, 2, 3],
             )
             σ = pm.HalfNormal("σ", 10, shape=n_components)
             weights = pm.Dirichlet("w", np.ones(n_components))
@@ -684,7 +684,7 @@ def __init__(self, w, comp_dists, mixture_axis=-1, *args, **kwargs):
         self.mixture_axis = mixture_axis
         kwargs.setdefault("dtype", self.comp_dists.dtype)
 
-        # Compute the mode so we don't always have to pass a testval
+        # Compute the mode so we don't always have to pass a initval
         defaults = kwargs.pop("defaults", [])
         event_shape = self.comp_dists.shape[mixture_axis + 1 :]
         _w = at.shape_padleft(

Diff for: pymc3/distributions/multivariate.py

@@ -840,7 +840,7 @@ def logp(self, X):
         )
 
 
-def WishartBartlett(name, S, nu, is_cholesky=False, return_cholesky=False, testval=None):
+def WishartBartlett(name, S, nu, is_cholesky=False, return_cholesky=False, initval=None):
     R"""
     Bartlett decomposition of the Wishart distribution. As the Wishart
     distribution requires the matrix to be symmetric positive semi-definite
@@ -875,7 +875,7 @@ def WishartBartlett(name, S, nu, is_cholesky=False, return_cholesky=False, testv
         Input matrix S is already Cholesky decomposed as S.T * S
     return_cholesky: bool (default=False)
         Only return the Cholesky decomposed matrix.
-    testval: ndarray
+    initval: ndarray
         p x p positive definite matrix used to initialize
 
     Notes
@@ -894,21 +894,21 @@ def WishartBartlett(name, S, nu, is_cholesky=False, return_cholesky=False, testv
     n_diag = len(diag_idx[0])
     n_tril = len(tril_idx[0])
 
-    if testval is not None:
+    if initval is not None:
         # Inverse transform
-        testval = np.dot(np.dot(np.linalg.inv(L), testval), np.linalg.inv(L.T))
-        testval = linalg.cholesky(testval, lower=True)
-        diag_testval = testval[diag_idx] ** 2
-        tril_testval = testval[tril_idx]
+        initval = np.dot(np.dot(np.linalg.inv(L), initval), np.linalg.inv(L.T))
+        initval = linalg.cholesky(initval, lower=True)
+        diag_testval = initval[diag_idx] ** 2
+        tril_testval = initval[tril_idx]
     else:
         diag_testval = None
         tril_testval = None
 
     c = at.sqrt(
-        ChiSquared("%s_c" % name, nu - np.arange(2, 2 + n_diag), shape=n_diag, testval=diag_testval)
+        ChiSquared("%s_c" % name, nu - np.arange(2, 2 + n_diag), shape=n_diag, initval=diag_testval)
     )
     pm._log.info("Added new variable %s_c to model diagonal of Wishart." % name)
-    z = Normal("%s_z" % name, 0.0, 1.0, shape=n_tril, testval=tril_testval)
+    z = Normal("%s_z" % name, 0.0, 1.0, shape=n_tril, initval=tril_testval)
     pm._log.info("Added new variable %s_z to model off-diagonals of Wishart." % name)
     # Construct A matrix
     A = at.zeros(S.shape, dtype=np.float32)

Diff for: pymc3/tests/models.py

@@ -30,7 +30,7 @@ def simple_model():
     mu = -2.1
     tau = 1.3
     with Model() as model:
-        Normal("x", mu, tau=tau, size=2, testval=floatX_array([0.1, 0.1]))
+        Normal("x", mu, tau=tau, size=2, initval=floatX_array([0.1, 0.1]))
 
     return model.initial_point, model, (mu, tau ** -0.5)
 
@@ -39,7 +39,7 @@ def simple_categorical():
     p = floatX_array([0.1, 0.2, 0.3, 0.4])
     v = floatX_array([0.0, 1.0, 2.0, 3.0])
     with Model() as model:
-        Categorical("x", p, size=3, testval=[1, 2, 3])
+        Categorical("x", p, size=3, initval=[1, 2, 3])
 
     mu = np.dot(p, v)
     var = np.dot(p, (v - mu) ** 2)
@@ -50,7 +50,7 @@ def multidimensional_model():
     mu = -2.1
     tau = 1.3
     with Model() as model:
-        Normal("x", mu, tau=tau, size=(3, 2), testval=0.1 * np.ones((3, 2)))
+        Normal("x", mu, tau=tau, size=(3, 2), initval=0.1 * np.ones((3, 2)))
 
     return model.initial_point, model, (mu, tau ** -0.5)
 
@@ -81,7 +81,7 @@ def simple_2model():
     tau = 1.3
     p = 0.4
     with Model() as model:
-        x = pm.Normal("x", mu, tau=tau, testval=0.1)
+        x = pm.Normal("x", mu, tau=tau, initval=0.1)
         pm.Deterministic("logx", at.log(x))
         pm.Bernoulli("y", p)
     return model.initial_point, model
@@ -91,7 +91,7 @@ def simple_2model_continuous():
     mu = -2.1
     tau = 1.3
     with Model() as model:
-        x = pm.Normal("x", mu, tau=tau, testval=0.1)
+        x = pm.Normal("x", mu, tau=tau, initval=0.1)
         pm.Deterministic("logx", at.log(x))
         pm.Beta("y", alpha=1, beta=1, size=2)
     return model.initial_point, model
@@ -106,7 +106,7 @@ def mv_simple():
             "x",
             at.constant(mu),
             tau=at.constant(tau),
-            testval=floatX_array([0.1, 1.0, 0.8]),
+            initval=floatX_array([0.1, 1.0, 0.8]),
         )
     H = tau
     C = np.linalg.inv(H)
@@ -122,7 +122,7 @@ def mv_simple_coarse():
             "x",
             at.constant(mu),
             tau=at.constant(tau),
-            testval=floatX_array([0.1, 1.0, 0.8]),
+            initval=floatX_array([0.1, 1.0, 0.8]),
         )
     H = tau
     C = np.linalg.inv(H)
@@ -138,7 +138,7 @@ def mv_simple_very_coarse():
             "x",
             at.constant(mu),
             tau=at.constant(tau),
-            testval=floatX_array([0.1, 1.0, 0.8]),
+            initval=floatX_array([0.1, 1.0, 0.8]),
         )
     H = tau
     C = np.linalg.inv(H)
@@ -150,7 +150,7 @@ def mv_simple_discrete():
     n = 5
     p = floatX_array([0.15, 0.85])
     with pm.Model() as model:
-        pm.Multinomial("x", n, at.constant(p), testval=np.array([1, 4]))
+        pm.Multinomial("x", n, at.constant(p), initval=np.array([1, 4]))
         mu = n * p
         # covariance matrix
         C = np.zeros((d, d))

Diff for: pymc3/tests/test_distributions_timeseries.py

@@ -68,13 +68,13 @@ def test_AR_nd():
     beta_tp = np.random.randn(p, n)
     y_tp = np.random.randn(T, n)
     with Model() as t0:
-        beta = Normal("beta", 0.0, 1.0, shape=(p, n), testval=beta_tp)
-        AR("y", beta, sigma=1.0, shape=(T, n), testval=y_tp)
+        beta = Normal("beta", 0.0, 1.0, shape=(p, n), initval=beta_tp)
+        AR("y", beta, sigma=1.0, shape=(T, n), initval=y_tp)
 
     with Model() as t1:
-        beta = Normal("beta", 0.0, 1.0, shape=(p, n), testval=beta_tp)
+        beta = Normal("beta", 0.0, 1.0, shape=(p, n), initval=beta_tp)
         for i in range(n):
-            AR("y_%d" % i, beta[:, i], sigma=1.0, shape=T, testval=y_tp[:, i])
+            AR("y_%d" % i, beta[:, i], sigma=1.0, shape=T, initval=y_tp[:, i])
 
     np.testing.assert_allclose(t0.logp(t0.initial_point), t1.logp(t1.initial_point))
 
@@ -150,7 +150,7 @@ def test_linear():
     # build model
     with Model() as model:
         lamh = Flat("lamh")
-        xh = EulerMaruyama("xh", dt, sde, (lamh,), shape=N + 1, testval=x)
+        xh = EulerMaruyama("xh", dt, sde, (lamh,), shape=N + 1, initval=x)
         Normal("zh", mu=xh, sigma=sig2, observed=z)
     # invert
     with model:

Diff for: pymc3/tests/test_model.py

@@ -57,7 +57,7 @@ def __init__(self, mean=0, sigma=1, name="", model=None):
         super().__init__(name, model)
         self.register_rv(Normal.dist(mu=mean, sigma=sigma), "v1")
         Normal("v2", mu=mean, sigma=sigma)
-        Normal("v3", mu=mean, sigma=Normal("sd", mu=10, sigma=1, testval=1.0))
+        Normal("v3", mu=mean, sigma=Normal("sd", mu=10, sigma=1, initval=1.0))
         Deterministic("v3_sq", self.v3 ** 2)
         Potential("p1", at.constant(1))
 
@@ -462,7 +462,7 @@ def test_make_obs_var():
     fake_model = pm.Model()
     with fake_model:
         fake_distribution = pm.Normal.dist(mu=0, sigma=1)
-        # Create the testval attribute simply for the sake of model testing
+        # Create the initval attribute simply for the sake of model testing
         fake_distribution.name = input_name
 
     # Check function behavior using the various inputs

Diff for: pymc3/tests/test_sampling.py

@@ -387,7 +387,7 @@ def test_shared_named(self):
                 mu=np.atleast_2d(0),
                 tau=np.atleast_2d(1e20),
                 size=(1, 1),
-                testval=np.atleast_2d(0),
+                initval=np.atleast_2d(0),
             )
             theta = pm.Normal(
                 "theta", mu=at.dot(G_var, theta0), tau=np.atleast_2d(1e20), size=(1, 1)
@@ -403,7 +403,7 @@ def test_shared_unnamed(self):
                 mu=np.atleast_2d(0),
                 tau=np.atleast_2d(1e20),
                 size=(1, 1),
-                testval=np.atleast_2d(0),
+                initval=np.atleast_2d(0),
             )
             theta = pm.Normal(
                 "theta", mu=at.dot(G_var, theta0), tau=np.atleast_2d(1e20), size=(1, 1)
@@ -419,7 +419,7 @@ def test_constant_named(self):
                 mu=np.atleast_2d(0),
                 tau=np.atleast_2d(1e20),
                 size=(1, 1),
-                testval=np.atleast_2d(0),
+                initval=np.atleast_2d(0),
             )
             theta = pm.Normal(
                 "theta", mu=at.dot(G_var, theta0), tau=np.atleast_2d(1e20), size=(1, 1)
@@ -688,10 +688,10 @@ def test_deterministic_of_observed_modified_interface(self):
         meas_in_1 = pm.aesaraf.floatX(2 + 4 * rng.randn(100))
         meas_in_2 = pm.aesaraf.floatX(5 + 4 * rng.randn(100))
         with pm.Model(rng_seeder=rng) as model:
-            mu_in_1 = pm.Normal("mu_in_1", 0, 1, testval=0)
-            sigma_in_1 = pm.HalfNormal("sd_in_1", 1, testval=1)
-            mu_in_2 = pm.Normal("mu_in_2", 0, 1, testval=0)
-            sigma_in_2 = pm.HalfNormal("sd__in_2", 1, testval=1)
+            mu_in_1 = pm.Normal("mu_in_1", 0, 1, initval=0)
+            sigma_in_1 = pm.HalfNormal("sd_in_1", 1, initval=1)
+            mu_in_2 = pm.Normal("mu_in_2", 0, 1, initval=0)
+            sigma_in_2 = pm.HalfNormal("sd__in_2", 1, initval=1)
 
             in_1 = pm.Normal("in_1", mu_in_1, sigma_in_1, observed=meas_in_1)
             in_2 = pm.Normal("in_2", mu_in_2, sigma_in_2, observed=meas_in_2)
@@ -882,7 +882,7 @@ def _mocked_init_nuts(*args, **kwargs):
 
 
 @pytest.mark.parametrize(
-    "testval, jitter_max_retries, expectation",
+    "initval, jitter_max_retries, expectation",
     [
         (0, 0, pytest.raises(SamplingError)),
         (0, 1, pytest.raises(SamplingError)),
@@ -891,9 +891,9 @@ def _mocked_init_nuts(*args, **kwargs):
         (1, 0, does_not_raise()),
     ],
 )
-def test_init_jitter(testval, jitter_max_retries, expectation):
+def test_init_jitter(initval, jitter_max_retries, expectation):
     with pm.Model() as m:
-        pm.HalfNormal("x", transform=None, testval=testval)
+        pm.HalfNormal("x", transform=None, initval=initval)
 
     with expectation:
         # Starting value is negative (invalid) when np.random.rand returns 0 (jitter = -1)

Diff for: pymc3/tests/test_step.py

@@ -964,25 +964,25 @@ def test_multiple_samplers(self, caplog):
 
     def test_bad_init_nonparallel(self):
         with Model():
-            HalfNormal("a", sigma=1, testval=-1, transform=None)
+            HalfNormal("a", sigma=1, initval=-1, transform=None)
             with pytest.raises(SamplingError) as error:
                 sample(init=None, chains=1, random_seed=1)
             error.match("Initial evaluation")
 
     @pytest.mark.skipif(sys.version_info < (3, 6), reason="requires python3.6 or higher")
     def test_bad_init_parallel(self):
         with Model():
-            HalfNormal("a", sigma=1, testval=-1, transform=None)
+            HalfNormal("a", sigma=1, initval=-1, transform=None)
             with pytest.raises(SamplingError) as error:
                 sample(init=None, cores=2, random_seed=1)
             error.match("Initial evaluation")
 
     def test_linalg(self, caplog):
         with Model():
-            a = Normal("a", size=2, testval=floatX(np.zeros(2)))
+            a = Normal("a", size=2, initval=floatX(np.zeros(2)))
             a = at.switch(a > 0, np.inf, a)
             b = at.slinalg.solve(floatX(np.eye(2)), a)
-            Normal("c", mu=b, size=2, testval=floatX(np.r_[0.0, 0.0]))
+            Normal("c", mu=b, size=2, initval=floatX(np.r_[0.0, 0.0]))
             caplog.clear()
             trace = sample(20, init=None, tune=5, chains=2)
             warns = [msg.msg for msg in caplog.records]