Add click install to fix ci (#193)

* Add black

Author: canyon289

.pre-commit-config.yaml

@@ -8,7 +8,7 @@ repos:
   rev: 0.5.6
   hooks:
     - id: nbqa-black
-      additional_dependencies: [black==20.8b1]
+      additional_dependencies: [black==22.3.0]
       files: ^(Rethinking_2|BSM)/
     - id: nbqa-isort
       additional_dependencies: [isort==5.6.4]

BSM/Chapter_03_00_Bayesian_CLT.ipynb

@@ -54,7 +54,7 @@
     "    A = Y - 0.5\n",
     "    B = n - Y - 0.5\n",
     "    θ_MAP = A / (A + B)\n",
-    "    info = A / θ_MAP ** 2 + B / (1 - θ_MAP) ** 2\n",
+    "    info = A / θ_MAP**2 + B / (1 - θ_MAP) ** 2\n",
     "\n",
     "    post1 = stats.binom(n, θ).pmf(Y) * stats.beta(0.5, 0.5).pdf(θ)\n",
     "    post1 = post1 / np.sum(post1)\n",

BSM/Chapter_03_01_Gibbs_sampling_one_sample_t-test.ipynb

@@ -142,7 +142,7 @@
     "    # sample mu|s2,Y\n",
     "    MN = np.sum(Y) / (n + m)\n",
     "    VR = s2 / (n + m)\n",
-    "    mu = stats.norm(MN, VR ** 0.5).rvs(1)\n",
+    "    mu = stats.norm(MN, VR**0.5).rvs(1)\n",
     "\n",
     "    # sample s2|mu,Y\n",
     "    A = a + n / 2\n",
@@ -291,7 +291,7 @@
     }
    ],
    "source": [
-    "keep_s = keep_s2 ** 0.5\n",
+    "keep_s = keep_s2**0.5\n",
     "plt.hist(keep_s2)\n",
     "plt.xlabel(\"sigma\")\n",
     "plt.title(\"Marginal posterior\");"

BSM/Chapter_03_02_Gibbs_sampling_two_sample_t-test.ipynb

@@ -122,12 +122,12 @@
     "    # sample muY|muZ,s2,Y,Z\n",
     "    A = np.sum(Y) / s2 + mu_0 / s2_0\n",
     "    B = n / s2 + 1 / s2_0\n",
-    "    muY = stats.norm(A / B, 1 / B ** 0.5).rvs(1)[0]\n",
+    "    muY = stats.norm(A / B, 1 / B**0.5).rvs(1)[0]\n",
     "\n",
     "    # sample muZ|muY,s2,Y,Z\n",
     "    A = np.sum(Z) / s2 + mu_0 / s2_0\n",
     "    B = m / s2 + 1 / s2_0\n",
-    "    muZ = stats.norm(A / B, 1 / B ** 0.5).rvs(1)[0]\n",
+    "    muZ = stats.norm(A / B, 1 / B**0.5).rvs(1)[0]\n",
     "\n",
     "    # sample s2|muY,muZ,Y,Z\n",
     "    A = n / 2 + m / 2 + a\n",

BSM/Chapter_03_03_Gibbs_sampling_for_simple_linear_regression.ipynb

@@ -394,12 +394,12 @@
     "    # sample alpha\n",
     "    V = n / s2 + mu_0 / s2_0\n",
     "    M = np.sum(Y - X * β) / s2 + 1 / s2_0\n",
-    "    α = stats.norm(M / V, 1 / V ** 0.5).rvs(1)[0]\n",
+    "    α = stats.norm(M / V, 1 / V**0.5).rvs(1)[0]\n",
     "\n",
     "    # sample beta\n",
-    "    V = np.sum(X ** 2) / s2 + mu_0 / s2_0\n",
+    "    V = np.sum(X**2) / s2 + mu_0 / s2_0\n",
     "    M = np.sum(X * (Y - α)) / s2 + 1 / s2_0\n",
-    "    β = stats.norm(M / V, 1 / V ** 0.5).rvs(1)[0]\n",
+    "    β = stats.norm(M / V, 1 / V**0.5).rvs(1)[0]\n",
     "\n",
     "    # sample s2|mu,Y,Z\n",
     "    A = n / 2 + a\n",

BSM/Chapter_03_09_Simple_linear_regression_in_PyMC3.ipynb

@@ -79,7 +79,7 @@
     "with pm.Model() as model:\n",
     "    # Priors\n",
     "    τ = pm.Gamma(\"τ\", 0.1, 10)\n",
-    "    σ = pm.Deterministic(\"σ\", 1 / (τ ** 0.5))\n",
+    "    σ = pm.Deterministic(\"σ\", 1 / (τ**0.5))\n",
     "    # σ = pm.HalfNormal('σ', np.std(mass))\n",
     "    β1 = pm.Normal(\"β1\", 0, 1000)\n",
     "    β2 = pm.Normal(\"β2\", 0, 1000)\n",

Rethinking_2/Chp_04.ipynb

@@ -3062,7 +3062,7 @@
    ],
    "source": [
     "d[\"weight_std\"] = (d.weight - d.weight.mean()) / d.weight.std()\n",
-    "d[\"weight_std2\"] = d.weight_std ** 2\n",
+    "d[\"weight_std2\"] = d.weight_std**2\n",
     "\n",
     "with pm.Model() as m_4_5:\n",
     "    a = pm.Normal(\"a\", mu=178, sd=100)\n",
@@ -3329,7 +3329,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "weight_m = np.vstack((d.weight_std, d.weight_std ** 2, d.weight_std ** 3))"
+    "weight_m = np.vstack((d.weight_std, d.weight_std**2, d.weight_std**3))"
    ]
   },
   {

Rethinking_2/Chp_06.ipynb

@@ -1217,7 +1217,7 @@
     "\n",
     "\n",
     "def sim_coll(r=0.9):\n",
-    "    x = np.random.normal(loc=r * d[\"perc.fat\"], scale=np.sqrt((1 - r ** 2) * np.var(d[\"perc.fat\"])))\n",
+    "    x = np.random.normal(loc=r * d[\"perc.fat\"], scale=np.sqrt((1 - r**2) * np.var(d[\"perc.fat\"])))\n",
     "    _, cov = curve_fit(mv, (d[\"perc.fat\"], x), d[\"kcal.per.g\"])\n",
     "    return np.sqrt(np.diag(cov))[-1]\n",
     "\n",

Rethinking_2/Chp_09.ipynb

@@ -139,7 +139,7 @@
    ],
    "source": [
     "def rad_dist(Y):\n",
-    "    return np.sqrt(np.sum(Y ** 2))\n",
+    "    return np.sqrt(np.sum(Y**2))\n",
     "\n",
     "\n",
     "fig, ax = plt.subplots(1, 1, figsize=[7, 3])\n",
@@ -207,8 +207,8 @@
     "def calc_U_gradient(x, y, q, a=0, b=1, k=0, d=1):\n",
     "    muy, mux = q\n",
     "\n",
-    "    G1 = np.sum(y - muy) + (a - muy) / b ** 2  # dU/dmuy\n",
-    "    G2 = np.sum(x - mux) + (k - mux) / b ** 2  # dU/dmux\n",
+    "    G1 = np.sum(y - muy) + (a - muy) / b**2  # dU/dmuy\n",
+    "    G2 = np.sum(x - mux) + (k - mux) / b**2  # dU/dmux\n",
     "\n",
     "    return np.array([-G1, -G2])"
    ]
@@ -257,9 +257,9 @@
     "    p *= -1\n",
     "    # Evaluate potential and kinetic energies sat start and end of trajectory\n",
     "    current_U = U(x, y, current_q)\n",
-    "    current_K = np.sum(current_p ** 2) / 2\n",
+    "    current_K = np.sum(current_p**2) / 2\n",
     "    proposed_U = U(x, y, q)\n",
-    "    proposed_K = np.sum(p ** 2) / 2\n",
+    "    proposed_K = np.sum(p**2) / 2\n",
     "    # Accept or reject the state at end of trajectory, returning either\n",
     "    # the position at the end of the trajectory or the initial position\n",
     "    accept = False\n",

Rethinking_2/Chp_10.ipynb

@@ -233,7 +233,7 @@
    ],
    "source": [
     "p = 0.7\n",
-    "A = [(1 - p) ** 2, p * (1 - p), (1 - p) * p, p ** 2]\n",
+    "A = [(1 - p) ** 2, p * (1 - p), (1 - p) * p, p**2]\n",
     "A"
    ]
   },
@@ -299,9 +299,9 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "H = np.zeros(10 ** 5)\n",
-    "p = np.zeros((10 ** 5, 4))\n",
-    "for rep in range(10 ** 5):\n",
+    "H = np.zeros(10**5)\n",
+    "p = np.zeros((10**5, 4))\n",
+    "for rep in range(10**5):\n",
     "    h, p_ = sim_p()\n",
     "    H[rep] = h\n",
     "    p[rep] = p_"

Rethinking_2/Chp_14.ipynb

@@ -108,7 +108,7 @@
    ],
    "source": [
     "cov_ab = sigma_a * sigma_b * rho\n",
-    "Sigma = np.array([[sigma_a ** 2, cov_ab], [cov_ab, sigma_b ** 2]])\n",
+    "Sigma = np.array([[sigma_a**2, cov_ab], [cov_ab, sigma_b**2]])\n",
     "Sigma"
    ]
   },
@@ -4321,7 +4321,7 @@
     "# linear\n",
     "ax.plot(xrange, np.exp(-1 * xrange), \"k--\", label=\"linear\")\n",
     "# squared\n",
-    "ax.plot(xrange, np.exp(-1 * xrange ** 2), \"k\", label=\"squared\")\n",
+    "ax.plot(xrange, np.exp(-1 * xrange**2), \"k\", label=\"squared\")\n",
     "\n",
     "ax.set_xlabel(\"distance\")\n",
     "ax.set_ylabel(\"correlation\")\n",
@@ -4606,14 +4606,14 @@
     "\n",
     "    etasq = pm.Exponential(\"etasq\", 2.0)\n",
     "    ls_inv = pm.HalfNormal(\"ls_inv\", 2.0)\n",
-    "    rhosq = pm.Deterministic(\"rhosq\", 0.5 * ls_inv ** 2)\n",
+    "    rhosq = pm.Deterministic(\"rhosq\", 0.5 * ls_inv**2)\n",
     "\n",
     "    # Implementation with PyMC's GP module:\n",
     "    cov = etasq * pm.gp.cov.ExpQuad(input_dim=1, ls_inv=ls_inv)\n",
     "    gp = pm.gp.Latent(cov_func=cov)\n",
     "    k = gp.prior(\"k\", X=Dmat)\n",
     "\n",
-    "    lam = (a * P ** b / g) * tt.exp(k[society])\n",
+    "    lam = (a * P**b / g) * tt.exp(k[society])\n",
     "\n",
     "    T = pm.Poisson(\"total_tools\", lam, observed=total_tools)\n",
     "\n",
@@ -4984,7 +4984,7 @@
     "# compute posterior median covariance\n",
     "x_seq = np.linspace(0, 10, 100)\n",
     "post = idata_14_8.posterior.stack(sample=(\"chain\", \"draw\"))\n",
-    "pmcov_mu = post[\"etasq\"].median().values * np.exp(-post[\"rhosq\"].median().values * (x_seq ** 2))"
+    "pmcov_mu = post[\"etasq\"].median().values * np.exp(-post[\"rhosq\"].median().values * (x_seq**2))"
    ]
   },
   {
@@ -5016,7 +5016,7 @@
     "    x_seq,\n",
     "    (\n",
     "        post[\"etasq\"][::50].values[:, None]\n",
-    "        * np.exp(-post[\"rhosq\"][::50].values[:, None] * (x_seq ** 2))\n",
+    "        * np.exp(-post[\"rhosq\"][::50].values[:, None] * (x_seq**2))\n",
     "    ).T,\n",
     "    \"k\",\n",
     "    alpha=0.08,\n",
@@ -5254,7 +5254,7 @@
    "source": [
     "# convert to correlation matrix\n",
     "sigma_post = np.sqrt(np.diag(K))\n",
-    "Rho = (sigma_post ** -1) * K * (sigma_post ** -1)\n",
+    "Rho = (sigma_post**-1) * K * (sigma_post**-1)\n",
     "\n",
     "# add row/col names for convenience\n",
     "Rho = pd.DataFrame(\n",

Rethinking_2/Chp_16.ipynb

@@ -126,7 +126,7 @@
     "h_ppc = np.linspace(0, 1.5, 100)\n",
     "\n",
     "for k, p in zip(prior_checks[\"k\"], prior_checks[\"p\"]):\n",
-    "    w_ppc = np.pi * k * p ** 2 * h_ppc ** 3\n",
+    "    w_ppc = np.pi * k * p**2 * h_ppc**3\n",
     "    ax.plot(h_ppc, w_ppc, c=\"k\", alpha=0.4)\n",
     "\n",
     "ax.scatter(d.h, d.w, c=\"C0\", alpha=0.3)\n",
@@ -325,7 +325,7 @@
    "source": [
     "w_sim = pm.sample_posterior_predictive(trace_16_1, 200, m16_1)\n",
     "h_seq = np.linspace(0, d.h.max(), 30)\n",
-    "mu_mean = np.pi * (trace_16_1[\"k\"] * trace_16_1[\"p\"] ** 2).mean() * h_seq ** 3"
+    "mu_mean = np.pi * (trace_16_1[\"k\"] * trace_16_1[\"p\"] ** 2).mean() * h_seq**3"
    ]
   },
   {

Rethinking_2/End_of_chapter_problems/Chapter_7.ipynb

@@ -540,7 +540,7 @@
     "    b = pm.Normal(\"b\", 0, 0.5, shape=2)  # beta prior\n",
     "    sigma = pm.Lognormal(\"sigma\", 0, 1)\n",
     "    x = pm.Data(\"x\", Laffer.s_taxRate)\n",
-    "    mu = pm.Deterministic(\"mu\", a + b[0] * x + b[1] * x ** 2)\n",
+    "    mu = pm.Deterministic(\"mu\", a + b[0] * x + b[1] * x**2)\n",
     "    rev = pm.Normal(\"rev\", mu=mu, sd=sigma, observed=Laffer.s_taxRevenue)\n",
     "    second_sample = pm.sample(draws=500, chains=4, return_inferencedata=True)\n",
     "\n",
@@ -549,7 +549,7 @@
     "    b = pm.Normal(\"b\", 0, 0.5, shape=3)  # beta prior\n",
     "    sigma = pm.Lognormal(\"sigma\", 0, 1)\n",
     "    x = pm.Data(\"x\", Laffer.s_taxRate)\n",
-    "    mu = pm.Deterministic(\"mu\", a + b[0] * x + b[1] * x ** 2 + b[2] * x ** 3)\n",
+    "    mu = pm.Deterministic(\"mu\", a + b[0] * x + b[1] * x**2 + b[2] * x**3)\n",
     "    rev = pm.Normal(\"rev\", mu=mu, sd=sigma, observed=Laffer.s_taxRevenue)\n",
     "    third_sample = pm.sample(draws=500, chains=4, return_inferencedata=True)\n",
     "\n",
@@ -558,7 +558,7 @@
     "    b = pm.Normal(\"b\", 0, 0.5, shape=4)  # beta prior\n",
     "    sigma = pm.Lognormal(\"sigma\", 0, 1)\n",
     "    x = pm.Data(\"x\", Laffer.s_taxRate)\n",
-    "    mu = pm.Deterministic(\"mu\", a + b[0] * x + b[1] * x ** 2 + b[2] * x ** 3 + b[3] * x ** 4)\n",
+    "    mu = pm.Deterministic(\"mu\", a + b[0] * x + b[1] * x**2 + b[2] * x**3 + b[3] * x**4)\n",
     "    rev = pm.Normal(\"rev\", mu=mu, sd=sigma, observed=Laffer.s_taxRevenue)\n",
     "    fourth_sample = pm.sample(draws=500, chains=4, return_inferencedata=True)"
    ]
@@ -966,7 +966,7 @@
     "    b = pm.Normal(\"b\", 0, 1, shape=2)  # beta prior\n",
     "    sigma = pm.Lognormal(\"sigma\", 0, 1)\n",
     "    x = pm.Data(\"x\", Laffer.s_taxRate)\n",
-    "    mu = pm.Deterministic(\"mu\", a + b[0] * x + b[1] * x ** 2)\n",
+    "    mu = pm.Deterministic(\"mu\", a + b[0] * x + b[1] * x**2)\n",
     "    rev = pm.Normal(\"rev\", mu=mu, sd=sigma, observed=Laffer.s_taxRevenue)\n",
     "    second_sample = pm.sample(draws=500, chains=4, return_inferencedata=True)\n",
     "\n",
@@ -975,7 +975,7 @@
     "    b = pm.Normal(\"b\", 0, 1, shape=3)  # beta prior\n",
     "    sigma = pm.Lognormal(\"sigma\", 0, 1)\n",
     "    x = pm.Data(\"x\", Laffer.s_taxRate)\n",
-    "    mu = pm.Deterministic(\"mu\", a + b[0] * x + b[1] * x ** 2 + b[2] * x ** 3)\n",
+    "    mu = pm.Deterministic(\"mu\", a + b[0] * x + b[1] * x**2 + b[2] * x**3)\n",
     "    rev = pm.Normal(\"rev\", mu=mu, sd=sigma, observed=Laffer.s_taxRevenue)\n",
     "    third_sample = pm.sample(draws=500, chains=4, return_inferencedata=True)\n",
     "\n",
@@ -984,7 +984,7 @@
     "    b = pm.Normal(\"b\", 0, 1, shape=4)  # beta prior\n",
     "    sigma = pm.Lognormal(\"sigma\", 0, 1)\n",
     "    x = pm.Data(\"x\", Laffer.s_taxRate)\n",
-    "    mu = pm.Deterministic(\"mu\", a + b[0] * x + b[1] * x ** 2 + b[2] * x ** 3 + b[3] * x ** 4)\n",
+    "    mu = pm.Deterministic(\"mu\", a + b[0] * x + b[1] * x**2 + b[2] * x**3 + b[3] * x**4)\n",
     "    rev = pm.Normal(\"rev\", mu=mu, sd=sigma, observed=Laffer.s_taxRevenue)\n",
     "    fourth_sample = pm.sample(draws=500, chains=4, return_inferencedata=True)"
    ]
@@ -1899,7 +1899,7 @@
     "    b = pm.Normal(\"b\", 0, 0.5, shape=2)\n",
     "    sigma = pm.Lognormal(\"sigma\", 0, 1)\n",
     "    x = pm.Data(\"x\", Laffer.s_taxRate)\n",
-    "    mu = pm.Deterministic(\"mu\", a + b[0] * x + b[1] * x ** 2)\n",
+    "    mu = pm.Deterministic(\"mu\", a + b[0] * x + b[1] * x**2)\n",
     "    rev = pm.StudentT(\"rev\", 2, mu=mu, sd=sigma, observed=Laffer.s_taxRevenue)\n",
     "    robust_second_sample = pm.sample(draws=500, chains=4, return_inferencedata=True)"
    ]