https://github.com/otrecoding/optimaltransportdataintegration.jl

Data integration using optimal transport theory
https://github.com/otrecoding/optimaltransportdataintegration.jl

data-integration julia-language optimal-transport

Last synced: 4 days ago
JSON representation

Data integration using optimal transport theory

• Host: GitHub
• URL: https://github.com/otrecoding/optimaltransportdataintegration.jl
• Owner: otrecoding
• License: mit
• Created: 2024-05-17T15:25:28.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2025-03-11T10:27:50.000Z (7 months ago)
• Last Synced: 2025-03-11T11:32:37.746Z (7 months ago)
• Topics: data-integration, julia-language, optimal-transport
• Language: Julia
• Homepage: https://otrecoding.github.io/OptimalTransportDataIntegration.jl/
• Size: 364 KB
• Stars: 0
• Watchers: 2
• Forks: 0
• Open Issues: 2
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# OptimalTransportDataIntegration.jl

[![CI](https://github.com/otrecoding/OptimalTransportDataIntegration.jl/actions/workflows/ci.yml/badge.svg)](https://github.com/otrecoding/OptimalTransportDataIntegration.jl/actions/workflows/ci.yml)

[![codecov](https://codecov.io/gh/otrecoding/OptimalTransportDataIntegration.jl/branch/master/graph/badge.svg)](https://codecov.io/gh/otrecoding/OptimalTransportDataIntegration.jl)

[![](https://img.shields.io/badge/docs-dev-blue.svg)](https://otrecoding.github.io/OptimalTransportDataIntegration.jl/dev)

This package implements a statistical matching strategy based on

optimal transport theory to integrate different data sources.

These data sources are related to the same target population, which

share a subset of covariates which each data source has its own

distinct subset of variables. After recoding you'll get a unique

data set in which all the variables, coming from the different

sources, are jointly available. 

This package is derived from

[OTRecod.jl](https://github.com/otrecoding/OTRecod.jl) where joint

distribution of shared and distinct variables is transported within

the data sources. Here the method also transports the distribution

of shared and distinct variables and estimates a function to predict

the missing variables.

## Installation

The package runs on julia 1.1 and above.

In a Julia session switch to `pkg>` mode to add the package:

```julia

julia>] # switch to pkg> mode

pkg> add https://github.com/otrecoding/OptimalTransportDataIntegration.jl

```

To run an example 

```julia

using OptimalTransportDataIntegration # import the package

params = DataParameters()  # Create the parameters set

rng = DiscreteDataGenerator(params)  # Create the random generator

data = generate( rng ) # Generate a dataset 

result = otrecod( data, JointOTBetweenBases() ) # Perform the statistical matching 

println(accuracy(result))  # Print accuracies on each distinct variables and the total accuracy.

```

It is possible tu use continuous explanatory variables by using

```

rng = ContinuousDataGenerator(params)

```

outcomes are always categorical, Y outcome levels are 1:4 and Z outcome levels are 1:3.