tensorflow / tensorflow/probability
Feature Request: Batched Update of RunningCentralMoments
Nobody has claimed this yet.
- Dominant language
- Jupyter Notebook
- Stars
- 4.4k
- Forks
- 1.1k
- PR merge metrics
- No merged PRs in 30d
Description
Adding new samples using RunningCentralMoments updates the exponentiated residuals and adjustment terms for each new individual sample, making it very slow to add a large number of new samples.
Pebay's formula(s) (see Proposition 2.1) allow for batched /partitioned updating of the central moments, similar to the current implementation of RunningMean and RunningVariance. Is there any reason that the batched version isn't implemented?
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
Locate the RunningCentralMoments implementation and compare it with the batched updating used by RunningMean and RunningVariance. Read Pebay's Proposition 2.1 to understand the partitioned formulas. Done means large batches can update the central moments without processing every sample individually, with behavior covered by the project's relevant tests.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- tensorflow
- Domain
- data
- Issue type
- Feature
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 38/100