Assembling ensembling: An adventure in approaches across disciplines
Data files
Jul 15, 2025 version files 25.88 KB
-
BaseTable.csv
9.54 KB
-
README.md
4.18 KB
-
Shiny-Dashboard.Rmd
12.15 KB
Abstract
When we think of model ensembling or ensemble modeling, there are many possibilities that come to mind in different disciplines. For example, one might think of a set of descriptions of a phenomenon in the world, perhaps a time series or a snapshot of multivariate space, and perhaps that set is comprised of data-independent descriptions, or perhaps it is quite intentionally fit to data, or even a suite of data sets with a common theme or intention. The very meaning of ‘ensemble’ - a collection together - conjures different ideas across and even within disciplines approaching phenomena. In this paper, we present a typology of the scope of these potential perspectives. It is not our goal to present a review of terms and concepts, nor is it to convince all disciplines to adopt a common suite of terms, which we view as futile. Rather, our goal is to disambiguate terms, concepts, and processes associated with ‘ensembles’ and ‘ensembling’ in order to facilitate communication, awareness, and possible adoption of tools across disciplines.
Dataset DOI: 10.5061/dryad.8gtht771c
Description of the data and file structure
About
When we think of model ensembling or ensemble modeling, there are many possibilities that come to mind in different disciplines. For example, one might think of a set of descriptions of a phenomenon in the world, perhaps a time series or a snapshot of multivariate space, and perhaps that set is comprised of data-independent descriptions, or perhaps it is quite intentionally fit to data, or even a suite of data sets with a common theme or intention. The very meaning of ‘ensemble’ - a collection together - conjures different ideas across and even within disciplines approaching phenomena. In this paper, we present a typology of the scope of these potential perspectives. It is not our goal to present a review of terms and concepts, nor is it to convince all disciplines to adopt a common suite of terms, which we view as futile. Rather, our goal is to disambiguate terms, concepts, and processes associated with ‘ensembles’ and ‘ensembling’ in order to facilitate communication, awareness, and possible adoption of tools across disciplines.
To help achieve this goal, we present a living compendium in dashboard form. The dashboard provides both a visual of the interconnectivity of ensembles across academic disciplines and a “living” table to facilitate the expansion of the static table found in the original paper.
A link to the preprint of the paper can be found at: https://arxiv.org/abs/2405.02599
The Dashboard
The dashboard includes two sections: (1) “The Big Picture” and (2) “Examples”. “The Big Picture” contains an interactive visualization of the main topics in the manuscript. Users can move, collapse, and expand the included nodes. The “Examples” page includes the living compendium, allowing users to add their own examples to the table. As can be seen in the sidebar of the page, there are five options that must be filled in to add an entry to the table: (1) Theme, (2) Discipline, (3) Ensemble Approach, (4) Manuscript Title, (5) Source.
When providing entries to the table, please include only one title and source per entry. However, if the manuscript falls under more than one theme and such is indicated upon submission, multiple rows will be created (i.e., one per theme). Please use the’; ’ to separate entries to indicate more than one discipline or ensemble approach for a given manuscript or source.
Once all fields have been completed and the “Update Table” button has been clicked, the entry will be submitted for approval by the site moderator. Once approved, the entry will be added to the table.
A link to the full interactive dashboard can be found at: https://collabdashboard.shinyapps.io/Shiny-Dashboard/#section-about
Files and variables
File: BaseTable.csv
Description: The base file used to create the dashboard shown in the R Shiny App.
Variables
- Theme: Underlying topic included within the manuscript of interest.
- Discipline: The discipline within which the manuscript of interest falls.
- Ensemble-Approach: The type of ensemble approach used within the manuscript of interest.
- Title: The title of the manuscript.
- Source: The URL or DOI of the manuscript of interest.
File: Shiny-Dashboard.Rmd
Description: The dashboard provides both a visual of the interconnectivity of ensembles across academic disciplines and a “living” table to facilitate the expansion of the static table found in the original paper. It requires both R and RStudio. A online version of the dashboard can be found at: https://collabdashboard.shinyapps.io/Shiny-Dashboard/#section-about.
Code/software
The dashboard requires both R and RStudio.
Access information
Other publicly accessible locations of the data: