This repository contains supplementary materials for the preprint paper, ***multiverse: Multiplexing Alternative Data Analyses in R Notebooks*** We provide link to the [multiverse](https://mucollective.github.io/multiverse/) R library repository on Github, which is going to be regularly updated and monitored. The repository (on OSF Storage) is structured as follows: - The file `multiverse_0.3.1.1.tar.gz` contains R library and can be used to install and run multiverse on your machine - To install, please execute the following command, replacing `<path_to_file>` with the appropriate path: `install.packages(<path_to_file>, repos = NULL, type="source")` - The directory labelled *hurricane-multiverse* contains an implementation of the hurricane example using `multiverse`. The enclosed R Markdown document can be used to run the analysis. The enclosed html file is a compiled version of this R Markdown document of the analysis which can be easily viewed in the browser. - The directory labelled *hurricane-rdfanalysis* contains an implementation of the hurricane example using [rdfanalysis](https://github.com/joachim-gassen/rdfanalysis). The enclosed R Markdown document can be used to run the analysis. The enclosed html file is a compiled version of this R Markdown document of the analysis which can be easily viewed in the browser. - rdfanalysis requires all the actual specification of the multiverse to be in dedicated template files. These can be found in the subdirectory `case-study-code`. - The directory labelled *hurricane-mverse* contains an implementation of the hurricane example using mverse. The enclosed R Markdown document can be used to run the analysis. The enclosed html file is a compiled version of this R Markdown document of the analysis which can be easily viewed in the browser. - The directory labelled *hurricane-boba* contains an implementation of the hurricane example using mverse. This was adapted from the implementation by [Liu et al.](https://github.com/uwdata/boba/tree/master/example/hurricane) which provides a detailed overview of their implementation. The directory contains: - the template.R file with the JSON specification that is passed into the boba compiler. - the dataset as a CSV file - `install.R` to import potentially missing R packages requires for execution - The directory labelled *user-interviews* contains materials used the informal, semi-structured formative evaluations. This directory contains: - `multiverse-pkg-desc.pdf` a brief overview of what a multiverse analysis is (as described in the literature) and the specific implementation aspects in the `multiverse` tool so that users get a quick idea of the functionality of the tool without having to go through extensive documentation within the constrained time frame of an interview. - In this semi-structured evaluation, we loosely followed a protocol outlined in `interview-guide.pdf` to get an idea of participants exeperience with R and implementing end-to-end data analysis. - In this semi-structured evaluation, we provided participants with data from a [paper by Li et al.](https://github.com/QishengLi/CHI2019_Reader_View) published in CHI 2019 as a single analysis implementation of a dataset. The metadata and the results of this study were provided to them, which is contained in the file `dataset-metadata.pdf` If you have any questions regarding this study or our paper, please contact Abhraneel at abhraneel@u.northwestern.edu