https://github.com/tpapp/batchols.jl

Fast likelihood-based OLS calculations with the same regressor matrix.
https://github.com/tpapp/batchols.jl

Last synced: 4 months ago
JSON representation

Fast likelihood-based OLS calculations with the same regressor matrix.

• Host: GitHub
• URL: https://github.com/tpapp/batchols.jl
• Owner: tpapp
• License: other
• Created: 2017-08-23T19:23:50.000Z (almost 9 years ago)
• Default Branch: master
• Last Pushed: 2020-12-20T09:13:11.000Z (over 5 years ago)
• Last Synced: 2026-01-20T20:04:08.197Z (5 months ago)
• Language: Julia
• Size: 6.84 KB
• Stars: 1
• Watchers: 1
• Forks: 1
• Open Issues: 3
• Metadata Files:
  • Readme: README.md
  • License: LICENSE.md

Awesome Lists containing this project

README

          
# BatchOLS

[![Project Status: WIP – Initial development is in progress, but there has not yet been a stable, usable release suitable for the public.](http://www.repostatus.org/badges/latest/wip.svg)](http://www.repostatus.org/#wip)

[![Build Status](https://travis-ci.org/tpapp/BatchOLS.jl.svg?branch=master)](https://travis-ci.org/tpapp/BatchOLS.jl)

[![Coverage Status](https://coveralls.io/repos/tpapp/BatchOLS.jl/badge.svg?branch=master&service=github)](https://coveralls.io/github/tpapp/BatchOLS.jl?branch=master)

[![codecov.io](http://codecov.io/github/tpapp/BatchOLS.jl/coverage.svg?branch=master)](http://codecov.io/github/tpapp/BatchOLS.jl?branch=master)

Maximum likelihood estimation and likelihood calculations for regressions of the form

```

y = X⋅β + ϵ

ϵ ∼ Normal(0, v), IID

```

where `β` are the coefficients and `v` is the variance of the error term.

The key features are

1. type stability, also for `ForwardDiff.Dual` numbers,

2. fast calculations for regressions using the same `X`.

Example:

```julia

import BatchOLS # no exported symbols

N = 100

K = 3

X = randn(N, K)

rhs = BatchOLS.RHS(X)

## maximum likelihood estimation

for _ in 1:100

    y = randn(N)

    β, v = BatchOLS.ML_βv(y, rhs)

end

## loglikelihood calculation

β′ = randn(K)

y = randn(N)

ℓ = BatchOLS.loglikelihood(y, rhs, β′, 1.0)

```

The code is fairly trivial, I put it in a package to allow rigorous automated testing after optimizations, especially for type inference.