Datavzrd: Rapid programming- and maintenance-free interactive visualization and communication of tabular data

Abstract

Tabular data, often scattered across multiple tables, is the primary output of data analyses in virtually all scientific fields. Exchange and communication of tabular data is therefore a central challenge. We present Datavzrd, a tool for creating portable, visually rich, interactive reports from tabular data in any kind of scientific discipline. Datavzrd unifies the strengths of currently common generic approaches for interactive visualization like R Shiny with the portability, ease of use and sustainability of plain spreadsheets. The generated reports do not require the maintenance of a web server nor the installation of specialized software for viewing and can simply be attached to emails, shared via cloud services, or serve as manuscript supplements. They can be specified without requiring imperative programming, thereby enabling rapid development and offering accessibility for non-computational scientists, unlocking the look and feel of dedicated manually crafted web applications without the maintenance and development burden. Datavzrd reports scale from small tables to thousands or millions of rows and offer the ability to link multiple related tables, allowing to jump between corresponding rows or hierarchically explore growing levels of detail.

Fig 1. Timeline comparison.

Comparison of the work items needed for publication and communication and the interaction capabilities of Datavzrd compared to above mentioned alternative approaches. Time per step is meant as a relative approximation of the true times, which may of course vary depending on the actual kind of data that shall be handled. We postulate that, for example, simple configuration requires less time than backend or frontend implementation and running a command line tool like Datavzrd is less effort than deploying a web service on a given infrastructure. Finally, the required effort for Datavzrd ends with the publication of the generated static HTML files, whereas server based approaches require continuous maintenance of their deployment in order to remain secure, operable, and accessible.

Fig 2. Screenshot of a Datavzrd report with annotated visual elements and controls.

The underlying example dataset entails genomic variants along with various scores and predictions. Gene names and coordinates have been altered to de-identify the data. Source: https://github.com/snakemake-workflows/dna-seq-varlociraptor. An interactive version of this report is available under https://datavzrd.github.io/example-molecular-tumor-board.

Fig 3. Example Datavzrd view of an astronomy spreadsheet supplement of Barraud et al. [22].

Boolean columns have been rendered with Datavzrd’s boolean spell [56], numeric columns have been rendered as heatmaps. The squarish link buttons at the far right of each row (generated by Datavzrd’s dataset linking functionality, see section Interactivity and visuals) allow to jump to corresponding rows in other views.

Fig 4. Datavzrd boolean spell and its application to a gene table.

Left: Datavzrd spell for displaying boolean values of the column highly_variable of the report to the right. The true_value and false_value parameters define which values are rendered into a plus sign and a minus sign. Right: Example Datavzrd view of a bioinformatics spreadsheet supplement of Klein et al. [20]. Gene names have been rendered as links to a public database, boolean values have been rendered using Datavzrd’s boolean spell [56] (see Community driven extensibility), other columns have been rendered as heatmaps with categorical or linear color scale.

Fig 5. Example Datavzrd view of a social science spreadsheet supplement of Alayón-Gamboa et al. [21].

Categorical columns have been rendered as heatmaps with a categorical color scale. The squarish link buttons at the far right of each row (generated by Datavzrd’s dataset linking functionality, see section Interactivity and visuals) allow to jump to corresponding rows in other views.

Fig 6. Example Datavzrd view of an anthropology spreadsheet supplement of L. G. Sanjuán et al. [19].

Numerical values have been rendered as tick plots or heatmaps. Categorical columns have been rendered as heatmaps with categorical color scale.

Fig 7. Storage usage of Datavzrd reports compared with Excel and raw input data.

Fig 8. Comparison of compression rates for increasing input sizes and different page sizes.

Fig 9. Comparison of search and filter mode (text input, filter brush or multi-select).

Fig 10. A Datavzrd dataset definition including different linkouts.

In this example, table-a creates a link to table-b where column gene-name of table-a and column gene of table-b match in value. With the gene details definition, table-a is linked to one of many tables existing for each value of gene-name.

Fig 11. Different pre-defined column visualizations of Datavzrd.

Left: corresponding YAML specification in the configuration file. Right: resulting column visualization. A: A Datavzrd tick plot definition with user-defined domain. B: A Datavzrd bar plot definition with user-defined domain and additional color domain. C: A Datavzrd heatmap for a column named Rated. D: Pill plot definition for a cell containing multiple values separated by any delimiter. E: YAML specification of a linkout to the NCBI gene database.