unpin black and reformat codebase (#1994)

* unpin black and reformat codebase

* reformat benchmarks

arviz/plots/backends/__init__.py

@@ -165,7 +165,7 @@ def create_layout(ax, force_layout=False):
         if subplot_order in ("square", "square_trimmed"):
             ax = [item for item in ax.ravel().tolist() if item is not None]
             n = int(np.ceil(len(ax) ** 0.5))
-            ax = ax + [None for _ in range(n ** 2 - len(ax))]
+            ax = ax + [None for _ in range(n**2 - len(ax))]
             ax = np.array(ax).reshape(n, n)
         ax = ax.tolist()
         if (subplot_order == "square_trimmed") and any(

arviz/plots/backends/bokeh/autocorrplot.py

@@ -76,7 +76,7 @@ def plot_autocorr(
         x_prime = x
         if combined:
             x_prime = x.flatten()
-        c_i = 1.96 / x_prime.size ** 0.5
+        c_i = 1.96 / x_prime.size**0.5
         y = autocorr(x_prime)
 
         ax.add_layout(BoxAnnotation(bottom=-c_i, top=c_i, fill_color="gray"))

arviz/plots/backends/bokeh/elpdplot.py

@@ -73,7 +73,7 @@ def plot_elpd(
 
     else:
         max_plots = (
-            numvars ** 2 if rcParams["plot.max_subplots"] is None else rcParams["plot.max_subplots"]
+            numvars**2 if rcParams["plot.max_subplots"] is None else rcParams["plot.max_subplots"]
         )
         vars_to_plot = np.sum(np.arange(numvars).cumsum() < max_plots)
         if vars_to_plot < numvars:

arviz/plots/backends/bokeh/pairplot.py

@@ -133,7 +133,7 @@ def plot_pair(
 
     dpi = backend_kwargs.pop("dpi")
     max_plots = (
-        numvars ** 2 if rcParams["plot.max_subplots"] is None else rcParams["plot.max_subplots"]
+        numvars**2 if rcParams["plot.max_subplots"] is None else rcParams["plot.max_subplots"]
     )
     vars_to_plot = np.sum(np.arange(numvars).cumsum() < max_plots)
     if vars_to_plot < numvars:

arviz/plots/backends/matplotlib/autocorrplot.py

@@ -50,7 +50,7 @@ def plot_autocorr(
         x_prime = x
         if combined:
             x_prime = x.flatten()
-        c_i = 1.96 / x_prime.size ** 0.5
+        c_i = 1.96 / x_prime.size**0.5
         y = autocorr(x_prime)
 
         ax.fill_between([0, max_lag], -c_i, c_i, color="0.75")

arviz/plots/backends/matplotlib/elpdplot.py

@@ -55,7 +55,7 @@ def plot_elpd(
                 for coord, float_color in color_mapping.items()
             ]
             plot_kwargs.setdefault("cmap", cmap_name)
-            plot_kwargs.setdefault("s", markersize ** 2)
+            plot_kwargs.setdefault("s", markersize**2)
             plot_kwargs.setdefault("c", colors)
         else:
             legend = False
@@ -70,7 +70,7 @@ def plot_elpd(
         (figsize, ax_labelsize, titlesize, xt_labelsize, _, markersize) = _scale_fig_size(
             figsize, textsize, numvars - 1, numvars - 1
         )
-        plot_kwargs.setdefault("s", markersize ** 2)
+        plot_kwargs.setdefault("s", markersize**2)
         backend_kwargs.setdefault("figsize", figsize)
         backend_kwargs["squeeze"] = True
         if ax is None:
@@ -113,7 +113,7 @@ def plot_elpd(
 
     else:
         max_plots = (
-            numvars ** 2 if rcParams["plot.max_subplots"] is None else rcParams["plot.max_subplots"]
+            numvars**2 if rcParams["plot.max_subplots"] is None else rcParams["plot.max_subplots"]
         )
         vars_to_plot = np.sum(np.arange(numvars).cumsum() < max_plots)
         if vars_to_plot < numvars:
@@ -128,7 +128,7 @@ def plot_elpd(
         (figsize, ax_labelsize, titlesize, xt_labelsize, _, markersize) = _scale_fig_size(
             figsize, textsize, numvars - 2, numvars - 2
         )
-        plot_kwargs.setdefault("s", markersize ** 2)
+        plot_kwargs.setdefault("s", markersize**2)
 
         if ax is None:
             fig, ax = plt.subplots(

arviz/plots/backends/matplotlib/khatplot.py

@@ -69,7 +69,7 @@ def plot_khat(
     hlines_kwargs["color"] = vectorized_to_hex(hlines_kwargs["color"])
 
     if markersize is None:
-        markersize = scaled_markersize ** 2  # s in scatter plot mus be markersize square
+        markersize = scaled_markersize**2  # s in scatter plot mus be markersize square
         # for dots to have the same size
 
     kwargs = matplotlib_kwarg_dealiaser(kwargs, "scatter")

arviz/plots/backends/matplotlib/pairplot.py

@@ -222,7 +222,7 @@ def plot_pair(
         not_marginals = int(not marginals)
         num_subplot_cols = numvars - not_marginals
         max_plots = (
-            num_subplot_cols ** 2
+            num_subplot_cols**2
             if rcParams["plot.max_subplots"] is None
             else rcParams["plot.max_subplots"]
         )

arviz/plots/plot_utils.py

@@ -125,7 +125,7 @@ def in_bounds(val):
 
         if n_items <= max_cols:
             return 1, n_items
-        ideal = in_bounds(round(n_items ** 0.5))
+        ideal = in_bounds(round(n_items**0.5))
 
         for offset in (0, 1, -1, 2, -2):
             cols = in_bounds(ideal + offset)

arviz/stats/density_utils.py

@@ -71,7 +71,7 @@ def _bw_isj(x, grid_counts=None, x_std=None, x_range=None):
     a_sq = a_k[range(1, grid_len)] ** 2
 
     t = _root(_fixed_point, x_len, args=(x_len, k_sq, a_sq), x=x)
-    h = t ** 0.5 * x_range
+    h = t**0.5 * x_range
     return h
 
 
@@ -100,8 +100,8 @@ def _bw_taylor(x):
     """
     x_len = len(x)
     kappa = _kappa_mle(x)
-    num = 3 * x_len * kappa ** 2 * ive(2, 2 * kappa)
-    den = 4 * np.pi ** 0.5 * ive(0, kappa) ** 2
+    num = 3 * x_len * kappa**2 * ive(2, 2 * kappa)
+    den = 4 * np.pi**0.5 * ive(0, kappa) ** 2
     return (num / den) ** 0.4
 
 
@@ -181,11 +181,11 @@ def _a1inv(x):
     Returns the value k, such that a1inv(x) = k, i.e. a1(k) = x.
     """
     if 0 <= x < 0.53:
-        return 2 * x + x ** 3 + (5 * x ** 5) / 6
+        return 2 * x + x**3 + (5 * x**5) / 6
     elif x < 0.85:
         return -0.4 + 1.39 * x + 0.43 / (1 - x)
     else:
-        return 1 / (x ** 3 - 4 * x ** 2 + 3 * x)
+        return 1 / (x**3 - 4 * x**2 + 3 * x)
 
 
 def _kappa_mle(x):
@@ -235,18 +235,18 @@ def _fixed_point(t, N, k_sq, a_sq):
     a_sq = np.asfarray(a_sq, dtype=np.float64)
 
     l = 7
-    f = np.sum(np.power(k_sq, l) * a_sq * np.exp(-k_sq * np.pi ** 2 * t))
+    f = np.sum(np.power(k_sq, l) * a_sq * np.exp(-k_sq * np.pi**2 * t))
     f *= 0.5 * np.pi ** (2.0 * l)
 
     for j in np.arange(l - 1, 2 - 1, -1):
         c1 = (1 + 0.5 ** (j + 0.5)) / 3
         c2 = np.product(np.arange(1.0, 2 * j + 1, 2, dtype=np.float64))
         c2 /= (np.pi / 2) ** 0.5
         t_j = np.power((c1 * (c2 / (N * f))), (2.0 / (3.0 + 2.0 * j)))
-        f = np.sum(k_sq ** j * a_sq * np.exp(-k_sq * np.pi ** 2.0 * t_j))
+        f = np.sum(k_sq**j * a_sq * np.exp(-k_sq * np.pi**2.0 * t_j))
         f *= 0.5 * np.pi ** (2 * j)
 
-    out = t - (2 * N * np.pi ** 0.5 * f) ** (-0.4)
+    out = t - (2 * N * np.pi**0.5 * f) ** (-0.4)
     return out
 
 
@@ -777,13 +777,13 @@ def _kde_adaptive(x, bw, grid_edges, grid_counts, grid_len, bound_correction, **
             [bw_adj[grid_npad - 1 :: -1], bw_adj, bw_adj[grid_len : grid_len - grid_npad - 1 : -1]]
         )
         pdf_mat = (grid_padded - grid_padded[:, None]) / bw_adj[:, None]
-        pdf_mat = np.exp(-0.5 * pdf_mat ** 2) * grid_counts[:, None]
+        pdf_mat = np.exp(-0.5 * pdf_mat**2) * grid_counts[:, None]
         pdf_mat /= (2 * np.pi) ** 0.5 * bw_adj[:, None]
         pdf = np.sum(pdf_mat[:, grid_npad : grid_npad + grid_len], axis=0) / len(x)
 
     else:
         pdf_mat = (grid - grid[:, None]) / bw_adj[:, None]
-        pdf_mat = np.exp(-0.5 * pdf_mat ** 2) * grid_counts[:, None]
+        pdf_mat = np.exp(-0.5 * pdf_mat**2) * grid_counts[:, None]
         pdf_mat /= (2 * np.pi) ** 0.5 * bw_adj[:, None]
         pdf = np.sum(pdf_mat, axis=0) / len(x)
 
@@ -838,7 +838,7 @@ def _fast_kde_2d(x, y, gridsize=(128, 128), circular=False):
     std_devs = np.diag(cov) ** 0.5
     kern_nx, kern_ny = np.round(scotts_factor * 2 * np.pi * std_devs)
 
-    inv_cov = np.linalg.inv(cov * scotts_factor ** 2)
+    inv_cov = np.linalg.inv(cov * scotts_factor**2)
 
     x_x = np.arange(kern_nx) - kern_nx / 2
     y_y = np.arange(kern_ny) - kern_ny / 2
@@ -854,8 +854,8 @@ def _fast_kde_2d(x, y, gridsize=(128, 128), circular=False):
     grid = coo_matrix((weights, xyi), shape=(n_x, n_y)).toarray()
     grid = convolve2d(grid, kernel, mode="same", boundary=boundary)
 
-    norm_factor = np.linalg.det(2 * np.pi * cov * scotts_factor ** 2)
-    norm_factor = len_x * d_x * d_y * norm_factor ** 0.5
+    norm_factor = np.linalg.det(2 * np.pi * cov * scotts_factor**2)
+    norm_factor = len_x * d_x * d_y * norm_factor**0.5
 
     grid /= norm_factor
 

arviz/stats/diagnostics.py

@@ -735,7 +735,7 @@ def _ess_sd(ary, relative=False):
     if _not_valid(ary, shape_kwargs=dict(min_draws=4, min_chains=1)):
         return np.nan
     ary = _split_chains(ary)
-    return min(_ess(ary, relative=relative), _ess(ary ** 2, relative=relative))
+    return min(_ess(ary, relative=relative), _ess(ary**2, relative=relative))
 
 
 def _ess_quantile(ary, prob, relative=False):

arviz/stats/stats.py

@@ -929,7 +929,7 @@ def _gpdfit(ary):
     prior_bs = 3
     prior_k = 10
     n = len(ary)
-    m_est = 30 + int(n ** 0.5)
+    m_est = 30 + int(n**0.5)
 
     b_ary = 1 - np.sqrt(m_est / (np.arange(1, m_est + 1, dtype=float) - 0.5))
     b_ary /= prior_bs * ary[int(n / 4 + 0.5) - 1]

arviz/tests/base_tests/test_stats.py

@@ -163,7 +163,7 @@ def test_r2_score():
     y = np.random.normal(x, 1)
     y_pred = x + np.random.randn(300, 100)
     res = linregress(x, y)
-    assert_allclose(res.rvalue ** 2, r2_score(y, y_pred).r2, 2)
+    assert_allclose(res.rvalue**2, r2_score(y, y_pred).r2, 2)
 
 
 @pytest.mark.parametrize("method", ["stacking", "BB-pseudo-BMA", "pseudo-BMA"])
@@ -627,7 +627,7 @@ def test_loo_pit_multi_lik():
         posterior={"a": np.random.randn(4, 100)},
         posterior_predictive={"y": post_pred},
         observed_data={"y": obs},
-        log_likelihood={"y": -(post_pred ** 2), "decoy": np.zeros_like(post_pred)},
+        log_likelihood={"y": -(post_pred**2), "decoy": np.zeros_like(post_pred)},
     )
     loo_pit_data = loo_pit(idata, y="y")
     assert np.all((loo_pit_data >= 0) & (loo_pit_data <= 1))

arviz/tests/base_tests/test_utils.py

@@ -360,7 +360,7 @@ def test_find_hdi_contours(mean, cov, contour_sigma):
     for idx, sigma in enumerate(contour_sigma):
         contour_sp[idx] = prob_dist.pdf(mean + sigma * stdevs[0] * eigenvecs[0])
 
-    hdi_probs = 1 - np.exp(-0.5 * contour_sigma ** 2)
+    hdi_probs = 1 - np.exp(-0.5 * contour_sigma**2)
     contour_az = _find_hdi_contours(density, hdi_probs)
 
     np.testing.assert_allclose(contour_sp, contour_az, rtol=1e-2, atol=1e-4)

arviz/tests/external_tests/test_data_emcee.py

@@ -123,13 +123,13 @@ def test_slices_warning(self, data, slices):
             from_emcee(data.obj, slices=slices)
 
     def test_no_blobs_error(self):
-        sampler = emcee.EnsembleSampler(6, 1, lambda x: -(x ** 2))
+        sampler = emcee.EnsembleSampler(6, 1, lambda x: -(x**2))
         sampler.run_mcmc(np.random.normal(size=(6, 1)), 20)
         with pytest.raises(ValueError):
             from_emcee(sampler, blob_names=["inexistent"])
 
     def test_peculiar_blobs(self, data):
-        sampler = emcee.EnsembleSampler(6, 1, lambda x: (-(x ** 2), (np.random.normal(x), 3)))
+        sampler = emcee.EnsembleSampler(6, 1, lambda x: (-(x**2), (np.random.normal(x), 3)))
         sampler.run_mcmc(np.random.normal(size=(6, 1)), 20)
         inference_data = from_emcee(sampler, blob_names=["normal", "threes"])
         fails = check_multiple_attrs({"log_likelihood": ["normal", "threes"]}, inference_data)
@@ -139,7 +139,7 @@ def test_peculiar_blobs(self, data):
         assert not fails
 
     def test_single_blob(self):
-        sampler = emcee.EnsembleSampler(6, 1, lambda x: (-(x ** 2), 3))
+        sampler = emcee.EnsembleSampler(6, 1, lambda x: (-(x**2), 3))
         sampler.run_mcmc(np.random.normal(size=(6, 1)), 20)
         inference_data = from_emcee(sampler, blob_names=["blob"], blob_groups=["blob_group"])
         fails = check_multiple_attrs({"blob_group": ["blob"]}, inference_data)

arviz/tests/helpers.py

@@ -291,9 +291,9 @@ def _emcee_lnprior(theta):
     # Half-cauchy prior, hwhm=25
     if tau < 0:
         return -np.inf
-    prior_tau = -np.log(tau ** 2 + 25 ** 2)
+    prior_tau = -np.log(tau**2 + 25**2)
     prior_mu = -((mu / 10) ** 2)  # normal prior, loc=0, scale=10
-    prior_eta = -np.sum(eta ** 2)  # normal prior, loc=0, scale=1
+    prior_eta = -np.sum(eta**2)  # normal prior, loc=0, scale=1
     return prior_mu + prior_tau + prior_eta
 
 

asv_benchmarks/benchmarks/benchmarks.py

@@ -55,7 +55,7 @@ def time_circ_std(self, numba_flag):
 
 
 class Kde_1d:
-    params = [(True, False), (10 ** 5, 10 ** 6, 10 ** 7)]
+    params = [(True, False), (10**5, 10**6, 10**7)]
     param_names = ("Numba", "n")
 
     def setup(self, numba_flag, n):
@@ -70,7 +70,7 @@ def time_fast_kde_normal(self, numba_flag, n):
 
 
 class Fast_KDE_2d:
-    params = [(True, False), ((100, 10 ** 4), (10 ** 4, 100), (1000, 1000))]
+    params = [(True, False), ((100, 10**4), (10**4, 100), (1000, 1000))]
     param_names = ("Numba", "shape")
 
     def setup(self, numba_flag, shape):

requirements-dev.txt

@@ -6,7 +6,7 @@ pydocstyle
 pylint
 pytest
 pytest-cov
-black==21.12b0 ; python_version >= '3.6'
+black ; python_version >= '3.6'
 typing_copilot ; python_version >= '3.7'
 mypy<0.800
 cloudpickle<1.5.0

requirements-docs.txt

@@ -12,5 +12,5 @@ sphinx-notfound-page
 sphinx-copybutton
 bokeh
 sphinx_design
-sphinx-codeautolink @ git+https://github.com/felix-hilden/sphinx-codeautolink
+sphinx-codeautolink>=0.9.0
 jupyter-sphinx