{WIP] Add HalfCauchy distribution #17
Conversation
| shape=shape, minval=0., maxval=1., dtype=self.dtype, seed=seed) | ||
| return tf.abs(sampled * self.scale) | ||
|
|
||
| def _prob(self, x): |
There was a problem hiding this comment.
Don't we get all of that from the bijection applied to the Cauchy distribution?
There was a problem hiding this comment.
As @jvdillon mentioned in the tensorflow_probability issue, the AbsoluteValue solution resembles HalfCauchy with similar properties but isn't exactly that.
From what I understand, AbsoluteValue gives a modified scaled up version of Cauchy resulting in similar properties to HalfCauchy(with same scale as Cauchy described earlier) only on the positive side. This is demonstrated in this notebook.
PS: The notebook uses HalfCauchy defined using absolute value.
IMO Only way to have a proper HalfCauchy distribution is to define it from scratch like tensorflow_probability does with other distributions.
|
Close as TFP now has HalfCauchy implemented. |
No description provided.