Use repr_html

preliz/distributions/distributions.py

@@ -39,37 +39,42 @@ def __init__(self):
         self.is_frozen = False
         self.opt = None
 
-    def __repr__(self):
+    def _get_name(self):
+        """Return the display name for this distribution."""
         name = self.__class__.__name__
         if name in ["Truncated", "Censored", "Hurdle"]:
             name += self.dist.__class__.__name__
         if name == "Mixture":
-            name = (
-                "Mixture"
-                + "".join(dict.fromkeys(dist.__class__.__name__ for dist in self.dist))
-                + "\n"
+            name = "Mixture" + "".join(dict.fromkeys(dist.__class__.__name__ for dist in self.dist))
+        return name
+
+    def _get_description(self):
+        """Return a string of parameters, or empty string if not frozen."""
+        if not self.is_frozen:
+            return ""
+        return "".join(
+            (
+                f"{n}={v:.3g}, "
+                if np.isscalar(v) or np.ndim(v) == 0
+                else f"{n}=["
+                + "".join(f"{vi:.3g}, " for vi in np.atleast_1d(v)).strip(", ")
+                + "], "
             )
+            for n, v in zip(self.param_names, self.params)
+        ).strip(", ")
 
+    def __repr__(self):
+        name = self._get_name()
         if self.is_frozen:
-            if "Mixture" in name:
-                bolded_name = "\033[1m" + name.strip() + "\033[0m" + "\n"
-            else:
-                bolded_name = "\033[1m" + name + "\033[0m"
-
-            description = "".join(
-                (
-                    f"{n}={v:.3g}, "
-                    if np.isscalar(v) or np.ndim(v) == 0
-                    else f"{n}=["
-                    + "".join(f"{vi:.3g}, " for vi in np.atleast_1d(v)).strip(", ")
-                    + "], "
-                )
-                for n, v in zip(self.param_names, self.params)
-            ).strip(", ")
+            return f"{name}({self._get_description()})"
+        return name
 
-            return f"{bolded_name}({description})"
-        else:
-            return name
+    def _repr_html_(self):
+        name = self._get_name()
+        if self.is_frozen:
+            desc = self._get_description()
+            return f"<span style='font-weight:bold'>{name}</span><span'>({desc})</span>"
+        return f"<span style='font-weight:bold'>{name}</span>"
 
     @property
     def params_dict(self):

preliz/internal/plot_helper.py

@@ -311,13 +311,6 @@ def side_legend(legend, ax):
         ax.legend(loc="center left", bbox_to_anchor=(1, 0.5))
 
 
-def repr_to_matplotlib(distribution):
-    string = repr(distribution)
-    string = string.replace("\x1b[1m", r"$\bf{")
-    string = string.replace("\x1b[0m", "}$")
-    return string
-
-
 def get_moments(dist, moments):
     names = {
         "m": "μ",
@@ -727,7 +720,7 @@ def set_label(dist, legend, moments, ax):
         if isinstance(legend, str) and legend not in ["title", "legend"]:
             label = legend
         else:
-            label = repr_to_matplotlib(dist)
+            label = str(dist)
 
             if moments is not None:
                 label += get_moments(dist, moments)

preliz/internal/plot_helper_multivariate.py

@@ -6,7 +6,6 @@
 import numpy as np
 from matplotlib import tri
 
-from preliz.internal.plot_helper import repr_to_matplotlib
 from preliz.internal.special import gammaln
 
 
@@ -194,7 +193,7 @@ def plot_dirichlet(
                 ax.set_ylim(*ylim)
 
         if legend == "title":
-            fig.text(0.5, 1, repr_to_matplotlib(dist), ha="center", va="center")
+            fig.text(0.5, 1, dist, ha="center", va="center")
 
     elif dim == 3:
         dirichlet_ = DirichletOnSimplex(alpha)
@@ -203,7 +202,7 @@ def plot_dirichlet(
                 _, axes = plt.subplots(1, 1)
             dirichlet_.plot(ax=axes)
             if legend == "title":
-                axes.set_title(repr_to_matplotlib(dist))
+                axes.set_title(dist)
     else:
         raise ValueError("joint only works for Dirichlet of dim=3")
 
@@ -354,14 +353,14 @@ def plot_mvnormal(
             if xy_lim != "auto" and representation != "cdf":
                 ax.set_ylim(*ylim)
         if legend == "title":
-            fig.text(0.5, 1, repr_to_matplotlib(dist), ha="center", va="center")
+            fig.text(0.5, 1, dist, ha="center", va="center")
 
     elif dim == 2:
         if axes is None:
             _, axes = plt.subplots(1, 1)
         joint_normal(dist, axes)
         if legend == "title":
-            axes.set_title(repr_to_matplotlib(dist))
+            axes.set_title(dist)
     else:
         raise ValueError("joint only works for Multivariate Normal of dim=2")
 

preliz/internal/predictive_helper.py

@@ -1,7 +1,6 @@
 import numpy as np
 
 from preliz.internal.distribution_helper import get_distributions
-from preliz.internal.plot_helper import repr_to_matplotlib
 from preliz.unidimensional import mle
 
 
@@ -11,7 +10,7 @@ def back_fitting_ppa(model, subset, new_families=True):
 
     for name, dist in model.items():
         dist._fit_mle(subset[name])
-        string += f"{name} = {repr_to_matplotlib(dist)}\n"
+        string += f"{name} = {dist}\n"
 
     if new_families:
         string += "\nYour selection is consistent with the priors (new families):\n"
@@ -26,7 +25,7 @@ def back_fitting_ppa(model, subset, new_families=True):
             elif dist.kind == "discrete":
                 distributions = get_distributions(set([dist.__class__.__name__] + common_disc))
             idx, _ = mle(distributions, subset[name], plot=False)
-            string += f"{name} = {repr_to_matplotlib(distributions[idx[0]])}\n"
+            string += f"{name} = {distributions[idx[0]]}\n"
 
     return string, np.concatenate([dist.params for dist in model.values()])
 

preliz/tests/test_posterior_to_prior.py

@@ -19,11 +19,9 @@
 
 def test_p2p_pymc():
     posterior_to_prior(model, idata)
-    assert 'Gamma\x1b[0m("b", alpha=' in posterior_to_prior(model, idata, new_families="auto")
+    assert 'Gamma("b", alpha=' in posterior_to_prior(model, idata, new_families="auto")
     posterior_to_prior(model, idata, new_families=[LogNormal()])
-    assert 'Gamma\x1b[0m("b", mu=' in posterior_to_prior(
-        model, idata, new_families={"b": [Gamma(mu=0)]}
-    )
+    assert 'Gamma("b", mu=' in posterior_to_prior(model, idata, new_families={"b": [Gamma(mu=0)]})
 
 
 # Temporarily disabled bambi test