adding sampling functionality to linearregression model #16
Conversation
There was a problem hiding this comment.
This seems like a good addition @jattenberg! Thanks for adding it. Do you plan on adding a unittest for this new method? I think it might also be useful to update the corresponding Jupyter notebook to show how you might use this method.
| @@ -3,6 +3,10 @@ All notable changes to this project will be documented in this file. | |||
|
|
|||
| The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/). | |||
|
|
|||
| ## [1.1.4] - 2018-07-21 | |||
There was a problem hiding this comment.
Hi, I'm not sure if you're familiar with semantic versioning; the last number is reserved for bug fixes.
This is not a bug fix, so the version should be a minor version change, e.g. 1.2.0
| @@ -93,17 +93,17 @@ def fit(self, X, y, inference_type='advi', minibatch_size=None, inference_args=N | |||
|
|
|||
| return self | |||
|
|
|||
| def predict(self, X, return_std=False): | |||
|
|
|||
| def sample(self, X, samples=2000): | |||
There was a problem hiding this comment.
What do you think about calling this something like sample_posterior? I think it could help to be a bit more explicit.
| @@ -93,17 +93,17 @@ def fit(self, X, y, inference_type='advi', minibatch_size=None, inference_args=N | |||
|
|
|||
| return self | |||
|
|
|||
| def predict(self, X, return_std=False): | |||
|
|
|||
There was a problem hiding this comment.
Only need one blank line between methods per PEP8.
|
|
||
| Parameters | ||
| ---------- | ||
| X : numpy array, shape [n_samples, n_features] | ||
|
|
||
| return_std : Boolean flag of whether to return standard deviations with mean values. Defaults to False. | ||
| samples : number of draws to make for each point |
There was a problem hiding this comment.
Could you add in the default value to this docstring? Like I do in the predict method below with return_std.
|
|
||
| return_std : Boolean flag of whether to return standard deviations with mean values. Defaults to False. | ||
|
|
||
| samples: numberof draws to make for each input |
There was a problem hiding this comment.
Typo: number of. Please also add the default value.
|
If my understanding is correct, the see Bayesian point estimation: https://en.wikipedia.org/wiki/Point_estimation#Bayesian_point_estimation |
|
@jattenberg Are you still working on this? |
i wanted to be able to sample predictions