Skip to content

AsynchroNous Disk-based Representation of MassivE DAta: An R package aimed at replacing ff for storing large data objects.

Notifications You must be signed in to change notification settings

OHDSI/Andromeda

Repository files navigation

Andromeda

Build Status codecov.io CRAN_Status_Badge CRAN_Status_Badge

Andromeda is part of HADES.

Introduction

AsynchroNous Disk-based Representation of MassivE DAta (ANDROMEDA): An R package for storing large data objects. Andromeda allow storing data objects on a local drive, while still making it possible to manipulate the data in an efficient manner.

Features

  • Allows storage of data objects much larger than what can fit in memory.
  • Integrates with dplyr package for data manipulation.
  • Objects are stored in a temporary location on the local file system.
  • Ability to save and load the objects to a compressed file in a permanent location on the local file system.

Examples

library(Andromeda)
bigData <- andromeda()

# Add some 'big' data:
bigData$cars <- cars

# Manipulate using dplyr:
bigData$cars %>% filter(speed > 10) %>% count() %>% collect()
# # A tibble: 1 x 1
#       n
#   <int>
# 1    41

saveAndromeda(bigData, "bigData.zip")
close(bigData)

Technology

The Andromeda package is an R package wrapped around RSQLite.

System Requirements

Running the package requires R.

Installation

To install the latest development version, install from GitHub:

install.packages("devtools")
devtools::install_github("ohdsi/Andromeda")

User Documentation

Documentation can be found on the package website.

Support

Contributing

Read here how you can contribute to this package.

License

Andromeda is licensed under Apache License 2.0

Development

Andromeda is being developed in R Studio.

Development status

Beta. Use at your own risk.

About

AsynchroNous Disk-based Representation of MassivE DAta: An R package aimed at replacing ff for storing large data objects.

Topics

Resources

Stars

Watchers

Forks