I have discovered the esquisse package in R (link). It is described as “a ‘shiny’ gadget to create ‘ggplot2’ charts interactively with drag-and-drop to map your variables. You can quickly visualize your data accordingly to their type, export to ‘PNG’ or ‘PowerPoint’, and retrieve the code to reproduce the chart”.
Information about the package, authors and maintainers can be found on CRAN (link).
I have compiled a brief GitHub repository to share some resources for this introduction. I include the data.frame I used (link).
My first attempt to use esquisse functionality:
I found this package one of the most intuitive CRAN packages I have used. I do have some experience with ggplot2 and understand that I will need to return to it to provide further details. The Shiny format of esquisse really appeals to me. I appreciated the ease of drag and drop that enabled me to modify my visualisations without the need to code.
It will become my first look tool for data visualisation.
I have been looking at the #AFLW 2019 data. I took the opportunity to include some CRAN packages I have not used before.
The data (two csv files) and my code are in a GitHub repository (link). My code is very basic and reflects my own thinking out loud as a I learn more about R.
In the past, I have tended to bookmark R suggestions and yet never manage to return to them as the list gets longer. My new practice is to create an R file to explore packages or code that strike me as interesting.
I used patchwork (link) and ggforce (link) in addition to ggrepel (link) to look at the data in the context of ggplot2 (link).
I was particularly interested in how patchwork helped me combine a range of images.
These work really well as PDF A4 pages and I thought they would be helpful summaries to stimulate conversation.
Researchers have some important decisions to make about the ways they share their discoveries.
Back in 2017, I was struck by Biecek Przemysław and Marcin Kosiński’s discussion of the use of the R package archivist (link). They discussed the opportunities we have to enable auditable and replicable analysis. Two years earlier, Data Carpentry facilitated a Reproducible Research in R workshop (link).
This week ,two finds have sent me off thinking about the explicit sharing of research journeys and discoveries.
The first find was Stencila, an open source project, that aims to make reproducible research more accessible (link). I noted that “Stencila provides a set of open-source software components enabling researchers to enable reproducible research … using interactive source code”.
I found Stencila through a link to Giuliano Maciocci, Michael Aufreiter and Nokome Bentley’s (2019) paper Introducing eLife’s first computationally reproducible article (link). This exemplifies the potential of a Reproducible Document Stack approach to open sharing. Researchers can use their existing word processing and spreadsheet tools and can embed R and Python code blocks that can generate live interactive plots using the Plotly.js library. Stencila uses the Mini formula language (link).
A second find, thanks to a Stephen Downes’ alert, was Alice Meadows, Laurel Haak and Josh Brown’s (2019) discussion of persistent identifiers (link). They note persistent identifiers “for people (researchers), places (their organizations) and things (their research outputs and other contributions) are foundational elements in the overall research information infrastructure”.
Supporting research includes supporting the research information infrastructure: the tools and services that researchers use which enable them to spend more time doing research and less time managing it – as well as the virtual building blocks on which those tools and services depend, such as metadata, standards and, the topic of this article, persistent identifiers (PIDs).
Meadows, Alice, Laurel L. Haak, and Josh Brown. 2019. “Persistent Identifiers: The Building Blocks of the Research Information Infrastructure”. Insights 32 (1): 9. DOI: http://doi.org/10.1629/uksg.457
I have mentioned before that one of my founding ideas for the International Journal of Performance Analysis in sport was to enable a paper in any language (with an English abstract or summary) that shared openly video and data resources. At that time the platforms available did not permit open sharing.
This week has brought back those memories of a global community sharing research journeys. It must be profoundly exciting entering the research community now or transforming existing practices as we become much more transparent about these journeys.