user2016

The Vernacular Ecology Index (VEI) is a newly proposed assessment method for sustainable urban development. It is composed of five elements (energy, culture, systems, placeness and vernacular) that are indicative of the spirit of the real and illusory within the context of the urban ecosystem. Within the components there are integrated indicators that aim to reflect and measure the viability of the individual element. When synthesized with their counterparts the index indicates strengths and areas in need of improvement within the designated study subject. Most importantly the index acts as a visual illustration of ecological progress as it is a critical intention to involve communities in the process of ecological appraisal, or put simply ‘mutual interaction’. One of the primary purposes of the VEI tool is to establish networks of benchmark practice in order to stimulate feedback loops to complimentary regions, ultimately benefitting the broader bioregion. Applying the index to a number of projects in a stipulated locality effectively offers an aerial image of the urban ecosystem’s health that could potentially pinpoint ecological strengths and weaknesses of the identified region.

This talk will illustrate how VEI’s graphical information model is developed using R’s grammar of graphics, which allows clear representation of the five categories with the ability to establish a rating for each of the components. Trough the use of shiny, R enables interactive communications with users for imputing VEI’s assessment data that can be presented on Google Maps for building spatial aerial image of a regional assessment.

Interactive statistical graphics toolkits play an important role in the exploratory phase of a data analysis cycle. Web graphics are rarely used during this phase, and are commonly reserved solely for the presentation of findings, mainly due to a lack of tools for quick iteration. The R package ggplot2 is a popular tool for data visualization with an elegant framework allowing useRs to quickly iterate through many plots with a minimal amount of friction. The R package plotly converts ggplot2 graphics to a web-based version, adding automatic support for interactivity such as pan, zoom, and identification (i.e., tooltips). It also has support for more advanced interactive techniques, such as linked brushing, thanks to infrastructure provided by the R packages shiny and htmlwidgets. In this talk, I'll present numerous examples that demonstrate these techniques, and how to they can used to derive insights from data.

Data analyses often produce many different data visualizations. Keeping track of these plots is crucial for both correctness and reproducibility of analytic results. Analysts typically resort to direct use of filenames and paths to organize and label their plots. Unfortunately, such ad hoc approaches do not scale well to longer and more complex analyses. Furthermore, locating specific plots months or years after the fact, when the chosen naming scheme has likely been forgotten, can be time consuming and painful. We propose a system which automatically tracks visualizations and annotates them with meaningful, searchable metadata. Beyond the benefits to individual analysts, the ability to search through plots created by others to discover analyses relevant to a particular dataset or research question is a powerful tool for facilitating collaboration and advancing science within multi-analyst, multi-project research departments and the wider scientific community. We present the viztrackr framework, a tool for tracking, automatically annotating, discovering, and reproducing statistical plots created in the R statistical programming language.

The R package 'GGally' provides several amalgam plots that build on the basic 'ggplot2' plotting system. Functions produce multivariate plots like generalized scatterplot matrices, and parallel coordinate plots are provided, network plots, survival models, and glyph maps for spatiotemporal data. This new work introduces a function, ggduo, to produce generalized plots for two groups of variables (e.g. a matrix of X variables and a matrix of Y variables), as might be modeled by multivariate regression, canonical correlation analysis or even multivariate time series. It builds on the structure of the ggmatrix functions used to produce the generalized pairs plot. The new function will help analysts to look at their data to support better modeling.

The rbokeh package is an R interface to the Bokeh visualization library. The interface is designed to be simple but flexible, allowing the expressiveness required for rapid generation of ad hoc statistical plots, while providing simple yet useful interactive capabilities such as tooltips and zoom/pan with nearly no additional effort, and additional customized interactivity with a little more effort. In this talk I will introduce rbokeh and show examples of using it to create a wide variety of interactive displays, including Shiny applications and Trelliscope.