Remove note that probs are automatically rescaled

This concerns the documentation of the Categorical and Multinomial
distributions.  Probabilities are indeed rescaled, but only when they
are numeric, and with a warning to discourage users from that.
Hence this should not be in the documentation.

See issue #5978.

Author: Armavica

pymc/distributions/discrete.py

@@ -1268,8 +1268,7 @@ class Categorical(Discrete):
     Parameters
     ----------
     p : array of floats
-        p > 0 and the elements of p must sum to 1. They will be automatically
-        rescaled otherwise.
+        p > 0 and the elements of p must sum to 1.
     logit_p : float
         Alternative log odds for the probability of success.
     """

pymc/distributions/multivariate.py

@@ -516,7 +516,7 @@ class Multinomial(Discrete):
     p : tensor_like of float
         Probability of each one of the different outcomes (0 <= p <= 1). The number of
         categories is given by the length of the last axis. Elements are expected to sum
-        to 1 along the last axis, and they will be automatically rescaled otherwise.
+        to 1 along the last axis.
     """
     rv_op = multinomial