-
Notifications
You must be signed in to change notification settings - Fork 4
Home
Welcome to the raadtools wiki!
RAAD tools is a developing system for R users, consisting of a file collection shared by local groups and user tools to access those files.
- raadtools is an R package that reads a variety of remote sensing, environmental and model data sets
- bowerbird is an R package that builds the data set collection by synchronizing against online sources These currently include the following sites, with over 1M individual files totalling ~13 Tb, though the actual data sets downloaded depend on specific configuration locally.
The blueant package hosts the actual configuration for most of the files we maintain, the main ones are listed in its data summary list.
The package bowerbird can be used to configure a subset of these collections, or a subset within an individual source (e.g. only some date range), or to add new sources for synchronization. We maintain a collection of files on RDSI storage in Hobart, with the help of TPAC - and at the Antarctic Division we rsync with the TPAC/RDSI collection to provide a local copy for Kingston staff (required by security limitations at the AAD). The Marine predator group at IMAS, Ecosystem Modelling at ACE, and some individual ACE/AAD researchers have raadtools- servers linked to the single collection on RDSI, and Kingston staff can use R from desktop or server environments for the rsync collection. In total this services ~35 people, with probably about 10 quite heavy users. The current arrangement has the following pros and cons: PROS
-
users on our existing Nectar servers can use raadtools
-
external users can be added to our Nectar servers (running in the Tasmania nodes) CONS
-
no servers or users outside the IMAS/TPAC or Kingston AAD networks can use our file collection adding new servers to read the RDSI/TPAC collection is manual
-
user management is manual, provided by individual staff responsible for building the Nectar machines (i.e. Mike and Ben)
We want to
provide the tools in a systematic way to staff and students, via some kind of containerized vm-cloud mechanism provide a Shiny service for users, where shiny pages can be developed and dropped into a place where they have access to raadtools promote the system to other groups, it seems natural to reproduce the collection at other places around Australia so localized networks can access them much as we do in Hobart (I have no opinion about how this is done) begin developing matching Python tools for those users