small fixes and tests

Author: aloctavodia

docs/gallery_content.rst

@@ -282,6 +282,18 @@ Continuous Distributions
       +++
       Rice
 
+   .. grid-item-card::
+      :link: ./distributions/gallery/scaledinversechisquared.html
+      :text-align: center
+      :shadow: none
+      :class-card: example-gallery
+
+      .. image:: distributions/img/ScaledInverseChiSquared.png
+         :alt: ScaledInverseChiSquared
+
+      +++
+      Scaled-Inverse-Chi-Squared
+
    .. grid-item-card::
       :link: ./distributions/gallery/skewnormal.html
       :text-align: center

docs/gallery_tags.rst

@@ -33,6 +33,7 @@ These distributions are defined over a continuous range of values.
 - `Normal <./distributions/gallery/normal.html>`_
 - `Pareto <./distributions/gallery/pareto.html>`_
 - `Rice <./distributions/gallery/rice.html>`_
+- `Scaled Inverse Chi-Squared <./distributions/gallery/scaledinversechisquared.html>`_
 - `Skew-Normal <./distributions/gallery/skewnormal.html>`_
 - `Student's t <./distributions/gallery/students_t.html>`_
 - `Skew-Student's t <./distributions/gallery/skew_studentt.html>`_
@@ -106,6 +107,7 @@ Distributions with skewed (non‐symmetric) shapes include:
 - `Moyal <./distributions/gallery/moyal.html>`_
 - `Pareto <./distributions/gallery/pareto.html>`_
 - `Rice <./distributions/gallery/rice.html>`_
+- `Scaled Inverse Chi-Squared <./distributions/gallery/scaledinversechisquared.html>`_
 - `Skew-Normal <./distributions/gallery/skewnormal.html>`_
 - `Skew-Student's t <./distributions/gallery/skew_studentt.html>`_
 - `Weibull <./distributions/gallery/weibull.html>`_
@@ -186,6 +188,8 @@ Distributions supported on the non‐negative real numbers.
 - `Wald <./distributions/gallery/wald.html>`_
 - `Log-Logistic <./distributions/gallery/log_logistic.html>`_
 - `Logit-Normal <./distributions/gallery/logit_normal.html>`_
+- `Scaled Inverse Chi-Squared <./distributions/gallery/scaledinversechisquared.html>`_
+
 
 **Discrete:**
 - `Bernoulli <./distributions/gallery/bernoulli.html>`_
@@ -243,6 +247,7 @@ The following distributions are univariate (one-dimensional).
 - `Normal <./distributions/gallery/normal.html>`_
 - `Pareto <./distributions/gallery/pareto.html>`_
 - `Rice <./distributions/gallery/rice.html>`_
+- `Scaled Inverse Chi-Squared <./distributions/gallery/scaledinversechisquared.html>`_
 - `Skew-Normal <./distributions/gallery/skewnormal.html>`_
 - `Student's t <./distributions/gallery/students_t.html>`_
 - `Skew-Student's t <./distributions/gallery/skew_studentt.html>`_

preliz/distributions/categorical.py

@@ -4,7 +4,7 @@
 from preliz.distributions.distributions import Discrete
 from preliz.internal.distribution_helper import all_not_none, eps
 from preliz.internal.optimization import optimize_ml
-from preliz.internal.special import expit, logit, xlogx
+from preliz.internal.special import expit, logit, ppf_bounds_disc, xlogx
 
 
 class Categorical(Discrete):
@@ -64,12 +64,6 @@ def _from_logit_p(self, logit_p):
     def _to_logit_p(self, p):
         return logit(p)
 
-    def _get_frozen(self):
-        frozen = None
-        if all_not_none(self.params):
-            frozen = self.dist(n=1, p=self.p)
-        return frozen
-
     def _update(self, p):
         self.p = np.array(p)
         self._n = len(p)
@@ -152,7 +146,7 @@ def nb_cdf(x, p):
 
 def nb_ppf(q, p):
     cumsum = np.cumsum(p)
-    return np.searchsorted(cumsum, q)
+    return ppf_bounds_disc(np.searchsorted(cumsum, q, side="left"), q, 0, len(p) - 1).astype(int)
 
 
 @nb.njit(cache=True)

preliz/distributions/chi_squared.py

@@ -3,7 +3,7 @@
 from scipy.special import gammainc, gammaincinv
 
 from preliz.distributions.distributions import Continuous
-from preliz.internal.distribution_helper import all_not_none, eps
+from preliz.internal.distribution_helper import eps
 from preliz.internal.optimization import optimize_ml
 from preliz.internal.special import cdf_bounds, digamma, gammaln, ppf_bounds_cont, xlogy
 
@@ -56,12 +56,6 @@ def _parametrization(self, nu=None):
         if self.nu is not None:
             self._update(self.nu)
 
-    def _get_frozen(self):
-        frozen = None
-        if all_not_none(self.params):
-            frozen = self.dist(self.nu)
-        return frozen
-
     def _update(self, nu):
         self.nu = np.float64(nu)
         self.params = (self.nu,)

preliz/distributions/discrete_uniform.py

@@ -61,12 +61,6 @@ def _parametrization(self, lower=None, upper=None):
             self.lower = lower
             self.upper = upper
 
-    def _get_frozen(self):
-        frozen = None
-        if all_not_none(self.params):
-            frozen = self.dist(self.lower, self.upper + 1)
-        return frozen
-
     def _update(self, lower, upper):
         self.lower = np.floor(lower)
         self.upper = np.ceil(upper)

preliz/distributions/kumaraswamy.py

@@ -2,7 +2,7 @@
 import numpy as np
 
 from preliz.distributions.distributions import Continuous
-from preliz.internal.distribution_helper import all_not_none, eps
+from preliz.internal.distribution_helper import eps
 from preliz.internal.optimization import optimize_ml, optimize_moments
 from preliz.internal.special import beta, cdf_bounds, digamma, ppf_bounds_cont, xlog1py, xlogy
 
@@ -57,12 +57,6 @@ def _parametrization(self, a=None, b=None):
         if (a and b) is not None:
             self._update(a, b)
 
-    def _get_frozen(self):
-        frozen = None
-        if all_not_none(self.params):
-            frozen = self.dist(self.a, self.b)
-        return frozen
-
     def _update(self, a, b):
         self.a = np.float64(a)
         self.b = np.float64(b)

preliz/distributions/logitnormal.py

@@ -79,12 +79,6 @@ def _parametrization(self, mu=None, sigma=None, tau=None):
         if all_not_none(mu, sigma):
             self._update(mu, sigma)
 
-    def _get_frozen(self):
-        frozen = None
-        if all_not_none(self.params):
-            frozen = self.dist(self.mu, self.sigma)
-        return frozen
-
     def _update(self, mu, sigma):
         self.mu = np.float64(mu)
         self.sigma = np.float64(sigma)

preliz/distributions/zi_negativebinomial.py

@@ -46,7 +46,7 @@ class ZeroInflatedNegativeBinomial(Discrete):
     ========  ==========================
     Support   :math:`x \in \mathbb{N}_0`
     Mean      :math:`\psi\mu`
-    Var       .. math::
+    Variance .. math::
                   \psi \left(\frac{{\mu^2}}{{\alpha}}\right) +\
                   \psi \mu + \psi \mu^2 - \psi^2 \mu^2
     ========  ==========================

preliz/tests/test_non_scipy.py

@@ -3,7 +3,17 @@
 from scipy import integrate
 from scipy.stats import kurtosis, skew
 
-from preliz.distributions import DiscreteWeibull, ScaledInverseChiSquared
+from preliz.distributions import Categorical, DiscreteWeibull, ScaledInverseChiSquared
+
+
+def test_categorical():
+    p = [0.2, 0.5, 0.3]
+    q = [0.2, 0.7, 1]
+    dist = Categorical(p)
+    assert dist.p.sum() == 1.0
+    assert all(dist.pdf([0, 1, 2]) == p)
+    assert all(dist.cdf([0, 1, 2]) == np.cumsum(p))
+    assert all(dist.cdf(dist.ppf(q)) == q)
 
 
 def test_disc_weibull_vs_random():

preliz/unidimensional/maxent.py

@@ -70,7 +70,6 @@ def maxent(
         :context: close-figs
         :include-source: true
 
-        >>>
         >>> import preliz as pz
         >>> pz.style.use('preliz-doc')
         >>> pz.maxent(pz.Gamma(), 1, 8, 0.9)
@@ -82,7 +81,6 @@ def maxent(
         :context: close-figs
         :include-source: true
 
-        >>>
         >>> import preliz as pz
         >>> pz.style.use('preliz-doc')
         >>> pz.maxent(pz.HalfStudentT(nu=4), 0, 12, 0.9)