scalanlp/breeze

Optionally Encode Dimension in Types

Open

#526 opened on Apr 18, 2016

View on GitHub
 (16 comments) (0 reactions) (0 assignees)Scala (3,453 stars) (690 forks)batch import
help wantedproject

Description

One thing I found tedious is the runtime check the dimensionality match, for example, when writing a neural network, if you want to be safe, you need to check if the dimensionality of each layer matches, and when writing a single layer, you need to check if the multiple components of that layer have matching dimensions.

I am wondering if it's possible to encode the dimension in type so that such runtime check can be compiler time check? For example, I imagine it's possible to use Shapeless.Nat something like Matrix with D(_2, _3).

Perhaps this has been considered and deemed infeasible?

Contributor guide

Tech stack
scala
Domain
datamachine learning
Issue type
feature
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
5
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
over 1 week
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
stale
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
unclear
Prerequisites
Scalatype level programmingShapeless
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
10
Research direction
Investigate the feasibility of encoding matrix dimensions in Scala types using Shapeless Nat. Review the discussion in issue comments to understand previous considerations and blocked efforts. Consider implementing a proof of concept using type level natural numbers to enforce compile time dimension checks.
Optionally Encode Dimension in Types · scalanlp/breeze#526 | Good First Issue