[ENH] Refactor of BaseDistribution and descendants - generalised di…

…stribution param broadcasting in base class (#5176)

Mirror of
[sktime/skpro#21

Moves the `_get_bc_params` method from child distributions to the parent
distribtion class, `BaseDistribution`.

This implementation is quite simple and doesn't use `_tags` and still
means the child distributions have to call the `_get_bc_params` method.

sktime/proba/base.py

@@ -158,6 +158,43 @@ def _method_error_msg(self, method="this method", severity="warn", fill_in=None)
         else:
             return msg
 
+    def _get_bc_params(self, *args, dtype=None):
+        """Fully broadcast tuple of parameters given param shapes and index, columns.
+
+        Parameters
+        ----------
+        args : float, int, array of floats, or array of ints (1D or 2D)
+            Distribution parameters that are to be made broadcastable. If no positional
+            arguments are provided, all parameters of `self` are used except for `index`
+            and `columns`.
+        dtype : str, optional
+            broadcasted arrays are cast to all have datatype `dtype`. If None, then no
+            datatype casting is done.
+
+        Returns
+        -------
+        Tuple of float or integer arrays
+            Each element of the tuple represents a different broadcastable distribution
+            parameter.
+        """
+        number_of_params = len(args)
+        if number_of_params == 0:
+            # Handle case where no positional arguments are provided
+            params = self.get_params()
+            params.pop("index")
+            params.pop("columns")
+            args = tuple(params.values())
+            number_of_params = len(args)
+
+        if hasattr(self, "index") and self.index is not None:
+            args += (self.index.to_numpy().reshape(-1, 1),)
+        if hasattr(self, "columns") and self.columns is not None:
+            args += (self.columns.to_numpy(),)
+        bc = np.broadcast_arrays(*args)
+        if dtype is not None:
+            bc = [array.astype(dtype) for array in bc]
+        return bc[:number_of_params]
+
     def pdf(self, x):
         r"""Probability density function.
 

sktime/proba/normal.py

@@ -45,7 +45,7 @@ def __init__(self, mu, sigma, index=None, columns=None):
         # and broadcast of parameters.
         # move this functionality to the base class
         # 0.19.0?
-        self._mu, self._sigma = self._get_bc_params()
+        self._mu, self._sigma = self._get_bc_params(self.mu, self.sigma)
         shape = self._mu.shape
 
         if index is None:
@@ -56,16 +56,6 @@ def __init__(self, mu, sigma, index=None, columns=None):
 
         super().__init__(index=index, columns=columns)
 
-    def _get_bc_params(self):
-        """Fully broadcast parameters of self, given param shapes and index, columns."""
-        to_broadcast = [self.mu, self.sigma]
-        if hasattr(self, "index") and self.index is not None:
-            to_broadcast += [self.index.to_numpy().reshape(-1, 1)]
-        if hasattr(self, "columns") and self.columns is not None:
-            to_broadcast += [self.columns.to_numpy()]
-        bc = np.broadcast_arrays(*to_broadcast)
-        return bc[0], bc[1]
-
     def energy(self, x=None):
         r"""Energy of self, w.r.t. self or a constant frame x.
 

sktime/proba/t.py

@@ -44,7 +44,9 @@ def __init__(self, mu, sigma, df=1, index=None, columns=None):
         self.index = index
         self.columns = columns
 
-        self._mu, self._sigma, self._df = self._get_bc_params()
+        self._mu, self._sigma, self._df = self._get_bc_params(
+            self.mu, self.sigma, self.df
+        )
         shape = self._mu.shape
 
         if index is None:
@@ -55,16 +57,6 @@ def __init__(self, mu, sigma, df=1, index=None, columns=None):
 
         super().__init__(index=index, columns=columns)
 
-    def _get_bc_params(self):
-        """Fully broadcast parameters of self, given param shapes and index, columns."""
-        to_broadcast = [self.mu, self.sigma, self.df]
-        if hasattr(self, "index") and self.index is not None:
-            to_broadcast += [self.index.to_numpy().reshape(-1, 1)]
-        if hasattr(self, "columns") and self.columns is not None:
-            to_broadcast += [self.columns.to_numpy()]
-        bc = np.broadcast_arrays(*to_broadcast)
-        return bc[0], bc[1], bc[2]
-
     def mean(self):
         r"""Return expected value of the distribution.
 

sktime/proba/tests/test_base_default_methods.py

@@ -34,7 +34,7 @@ def __init__(self, mu, sigma, index=None, columns=None):
         self.index = index
         self.columns = columns
 
-        self._mu, self._sigma = self._get_bc_params()
+        self._mu, self._sigma = self._get_bc_params(self.mu, self.sigma)
         shape = self._mu.shape
 
         if index is None:

sktime/proba/tfp.py

@@ -52,7 +52,7 @@ def __init__(self, mu, sigma, index=None, columns=None):
         # and broadcast of parameters.
         # move this functionality to the base class
         # 0.19.0?
-        self._mu, self._sigma = self._get_bc_params()
+        self._mu, self._sigma = self._get_bc_params(self.mu, self.sigma, dtype="float")
         distr = tfd.Normal(loc=self._mu, scale=self._sigma)
         shape = self._mu.shape
 
@@ -64,16 +64,6 @@ def __init__(self, mu, sigma, index=None, columns=None):
 
         super().__init__(index=index, columns=columns, distr=distr)
 
-    def _get_bc_params(self):
-        """Fully broadcast parameters of self, given param shapes and index, columns."""
-        to_broadcast = [self.mu, self.sigma]
-        if hasattr(self, "index") and self.index is not None:
-            to_broadcast += [self.index.to_numpy().reshape(-1, 1)]
-        if hasattr(self, "columns") and self.columns is not None:
-            to_broadcast += [self.columns.to_numpy()]
-        bc = np.broadcast_arrays(*to_broadcast)
-        return np.array(bc[0], dtype="float"), np.array(bc[1], dtype="float")
-
     def energy(self, x=None):
         r"""Energy of self, w.r.t. self or a constant frame x.