tensorflow / tensorflow/probability
Easy repetition of distribution parameters
Nobody has claimed this yet.
- Dominant language
- Jupyter Notebook
- Stars
- 4.4k
- Forks
- 1.1k
- PR merge metrics
- No merged PRs in 30d
Description
Hi all,
I'm trying to figure out if there is an implemented, general way of taking a tfp.distribution, repeating certain batch members, and making a new distribution to sample from. The code below works, but gets cumbersome quickly when I need to do the repetition on many different types of distributions.
import tensorflow as tf
import tensorflow_probability as tfp
norm = tfp.distributions.Normal([[0,1],[2,3]], [[1,2],[3,4]])
### what if you want to repeat the first distribution multiple times?
norm_repeat = tfp.distributions.Normal(tf.gather(norm.loc, [0,0,0,1]), tf.gather(norm.scale, [0,0,0,1]))
norm_repeat.sample()
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start with the tfp.distributions.Normal example in the issue and review how distribution batch members are represented and sampled. Determine the general behavior needed for repeating selected batch members across distribution types; done means a reusable distribution-level approach is defined and its sampling behavior is verified.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- machine-learning
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 30/100