NOTE: Scroll down to get to the tutorial itself!
Updated November 2015 for Palladio 1.1. If you’d like to use this tutorial in the classroom, or if you want to alter it and make it your own, there’s a version on Github you can do whatever you want with.
Palladio, a product of Stanford’s Humanities + Design Lab, is a web-based visualization tool for complex humanities data. Think of Palladio as a sort of Swiss Army knife for humanities data. It’s one package that includes a number of tools, each of which allows you to get a different angle on the same data.
Palladio is relatively new and still under active development which means that you will almost certainly encounter bugs! Still, it’s a very useful tool for getting a handle on a complicated dataset.
When Might Palladio be the Right Tool for You?
You have structured data.
Here, “structured data” means “data in a spreadsheet”: categorized, sorted, and stored in an Excel document or some other kind of spreadsheet application.
You’re interested in time, space, and relationships.
That’s where Palladio excels: showing you how various entities are connected across time and space.
Your data has many attributes.
Palladio’s really good at helping you uncover relationships among disparate attributes over time and space for example, it can help you see that a diarist was especially interested in trees as he traveled through North Carolina, and especially interested in bats as he traveled through Arizona. Palladio allows you to drill down through your data using faceted browsing.
When Might Palladio Not be the Right Tool for You?
You have unstructured data.
If you’re trying to analyze a long text, like a poem or a novel, Palladio won’t help you much. You’ll want to look for text analysis tools, like Voyant (http://voyant-tools.org/).
You just want to count things.
If you just want to make relatively simple charts and graphs, like a bar or pie chart, Palladio is too much tool for you! Instead, try using Excel’s built-in functions, or check out tools like Plot.ly or Tableau.
You want to present an interactive visualization.
One big limitation of Palladio is that you can’t embed or share the visualizations you create, except in static form. So while Palladio can help you explore and understand your data, it’s not great for presentation, at least not yet. Instead, try Google Fusion Tables, ManyEyes, or Tableau.
You want to create complex, fine-tuned maps and networks graphs.
While Palladio can produce maps and network graphs, you can’t customize them to any great extent, and you can’t perform sophisticated network analysis, such as calculating various measures of centrality. Instead, you might consider more sophisticated mapping tools, such as CartoDB or ArcGIS, and more sophisticated network analysis tools, such as Gephi and Cytoscape.
You hate bugs.
Palladio is still a baby, and you will almost certainly encounter some bugs. If you prefer not to use unstable software, you might investigate Google Fusion Tables or Tableau.
With that out of the way, we’re almost ready to get started using Palladio. First, though, a quick note that this tutorial does not cover some important features of Palladio, specifically its ability to link multiple data tables together, its timespan feature, and a feature that allows you to use multiple basemaps. Perhaps these will be the subject of a later tutorial!
A word on the dataset we’ll use, which you can find here.
This is a spreadsheet that contains the metadata for a portion of the Charles Weever Cushman Collection of photographs, located at Indiana University. The full Cushman Collection contains more than 14,500 Kodachrome photographs, taken between 1938 and 1969. Indiana University’s archivists were forward-thinking enough to place this data on Github, which is how we’re able to use it.
In order to make this data a little easier to work with, I’ve limited this spreadsheet to photographs taken between 1938 and 1955. I’ve also removed the “End Date” field to prevent confusion, changed the format of the date field, and added geocoordinates so that we can map the data more easily. For a great introduction to how to do some of this data manipulation on your own data, see this handout, developed by Owen Stephens on behalf of the British Library, which explains how to use the data-cleaning application OpenRefine.
A reminder that Palladio is still under development, so it can be buggy and slow! Some tips:
- Work slowly. Wait for an option to finish loading before you click it again or click something else.
- Do not refresh the page. You’ll lose your work.
- On a related note: To start over, refresh the page.
- Clicking on the Palladio logo will bring you to the Palladio homepage, but it won’t erase your work.