I am pleased to announce the release of DH Press version 2.5! It features many enhancements to the existing toolkit, plus six new visualizations.
The Pinboard visualization functions similarly to the map view, but instead of using an underlying Cartesian map, project admins can pin content to a background image of their own choosing. We have found this to be an excellent workaround for using historical maps without having to georectify them – see these examples from Anne Whisnant’s Parks to the Side project: Linville Falls Pinboard, Rocky Knob Pinboard.
Instead of layering additional maps on top of the base background image, the Pinboard supports the use of SVG layering. Project admins can create and load SVG layers to DH Press to enhance the Pinboard — to annotate images, similar to Scalar’s image annotation functionality. In addition to using DH Press markers to pin objects to the Pinboard, SVGs function as a secondary layer of content, but one that is not tied to the actual data model of a given project. SVG layers can be used to provide additional visual prompts to users, or explain elements of the background image.
Enhanced Pinboard with Animation
It is also possible to add an animation to a Pinboard by tying the optional SVG layers to a media file. Site visitors can play the media file and see various SVG layers turn on/off in sync with the streaming media. This feature can be used to provide a guided tour through the Pinboard. For example, the SVG layers could be tied to a tutorial video filmed by the project team, and used as a way to explain aspects of the interactive visualization.
The animation functions similarly to the Transcript Widget. Elements of the SVG file are assigned to different chunks of the media file via a timestamped plain text transcript (the “animation script”). Currently only YouTube media files work. Note that users will only be able to see, and not hear, the file.
Users have been asking for a timeline view since DH Press was still just a prototype. With version 2.5, it is now possible to provide a temporal display of your data. Our timeline supports the visual display of date ranges (e.g. a decade or a century) and single days, both BCE and CE.
Moreover, the DH Press timeline supports what we call “fuzzy” or uncertain dates. Project admins can specify uncertainty in their data, and DH Press will display that uncertainty by fading out the fill color of the timeline entry (see this example).
Fuzzy dates are critical to making the timeline a useful visualization for humanists. Given our uncertainty about much of the past, it allows us to provide a sense of when things might have happened without misleading our users, in a capta-based approach to data visualization (Johanna Drucker, “Humanities Approaches to Graphical Display,” Digital Humanities Quarterly 5, no. 1 (2011): link).
Facet flows allow you to visualize the relationship between different attributes in your data. For example, this visualization from the How Do You Say It? project shows the relationship between the intended gender audience for a given document about interpersonal violence prevention (Attribute A), and the actual use of gendered grammar, or gendered pronouns determinative articles (Attribute B). The colored bands can be rearranged by clicking on the “alpha” or “size” links for either attribute. Attribute positions can be swapped by dragging the top down to the bottom and vice versa. Hovering over a colored band shows the percentage of relevant results in the data. Clicking on a colored band will bring up all relevant documents in the window below.
The facet browser allows users to explore the project by filtering on any number of attributes. Similar to many e-commerce websites (such as Amazon or Zappos), users can drill down into several attributes at the same time in order to find specific elements of the data (“faceted search”). Project admins may add as many attributes (motes) to this view as they like, though too many will hinder the user experience.
The final visualization is the tree view, which displays 1 to N relationships within your data. The most common usage of this is for genealogical family trees, though the tree view can be used for any sort of parent-child relationship in your data (see this non-familial example). There are three different tree views available:
- a Flat Tree
- a Radial Tree, where each successive generation is placed at a further radius from the center root node
- a Segmented Wheel, where each successive generation is a further ring-layer moving out from the center core
New Data Types
To support these new visualizations, we’ve added several new data types:
- x,y coordinates for Pinboards (similar to latitude,longitude pairs)
- pointer data for Tree Views
- dates and date ranges for Timelines
- YouTube data for the Transcript Widget (see below) and Pinboard Animations
Pointer data can also be used to create relationships between data points without using the Tree View. Project admin can specify a “primary key” in the project to mimic the functionality of a relational database. Setting a primary key tells DH Press to treat a particular attribute of your data (custom field) as the unique identifier, which can be used to connect data points.
N.B. these data types are customized exclusively for DH Press. They are not globally recognized data types (such as nominal, integer, and ratio data). They were created to aid project admins in creating and configuring their visualizations. Please consult our documentation for detailed specifications about how to set up your data.
Enhancements to Current Features
In addition to these new visualizations, Michael has added some important functionality to existing DH Press capabilities. Most importantly, he has enhanced the Audio/Transcript tool (now called the “Transcript Widget” in v. 2.5) to support streaming media from either SoundCloud or YouTube. Additionally, he has extended the tool to support a scrolling transcript without a streaming audio/video file. This can be used for manuscript transcription.
For users of older versions of DH Press, make sure to consult our 2.5 Documentation for small changes in the project configuration dashboard, particularly with respect to configuring the Transcript Widget. Users are strongly encouraged to follow the formatting specifications for transcripts in this documentation.
We now have the capability to export data out of DH Press. Project admins can create a CSV file directly from their project dashboard. This is particularly useful in cases where many changes were made to the data after import. This enables project teams to update their data in DH Press without having to go back and make the same changes in their original data files. Instead, they can simply export the data after those changes have been made.
Start Using DH Press 2.5
DH Press 2.5 can be downloaded from our GitHub repository. Please consult Michael’s documentation about configuring data and projects for 2.5.
Note that once you migrate your project to 2.5, you will not be able to revert back to an older version of DH Press. We anticipate future releases to support backwards compatibility, but cannot at this time due to significant changes in the plugin’s architecture. For full details about migrating from 2.0 to 2.5, please consult our documentation.
Special thanks to Michael Newton, DH Press Lead Developer, for all of his hard work getting us to version 2.5. Additional thanks go out to the kind folks at OASIS, who have proven invaluable in the process, and to Anne MacNeil, one of our 2014 DIL/IAH Faculty Fellows, who assisted with field testing version 2.5, and whose project inspired many of our new visualizations.