Experiments with HGIS in Tableau

[As part of my PhD dissertation, I’m working on a Historical Geographic Information Systems (HGIS) project which traces the movement of people, objects, and ideas during the nineteenth century gold rushes. Previous posts on my Gold Rush Flow-Mapping project here:

A Graduate Fellow Introduction

Maps and Mining: Getting Data Out of ArcMap & the Beginning of Flow Mapping

Data Drudgery & Beautiful Betas]

Experiments with HGIS in Tableau

Over the past few months I’ve been continuing to add entries to my “moved things” database. Although necessary for the overall success of my project, this kind of work hardly makes for thrilling blog posts.

As part of my neverending quest to find the Best Ever Software for Flow Mapping, I’ve been testing a new program: Tableau. Under the pressure of their 14 day free trial, I managed to churn out some awesome results.

Learning Curve

It took me about 6 hours of tutorials and general messing around in the software to start producing legible maps. This compares well with ArcGIS, which took months and several expensive workshops. However, Tableau had been sold to me as a user-friendly, quick, and easy alternative to Arc…and so the learning curve, though minimal, was frustrating.

Its also worth noting that Tableau’s tutorials are almost uniformly unhelpful. Most assumed a clean dataset, pre-formatted for compatibility with the program. Eventually I figured out that for the purposes of flow maps, this meant that 1) all lat and long measurements needed to be in the same column and 2) origin and destination needed to be next to each other in the table. My database does not look that way, and it took quite a lot of effort to figure out how to format it so that Tableau could “read” it.

Results

Eventually I got Tableau to start spitting out things I wanted. Here’s the flow map it created for me, which shows moved ideas, objects, and people to the Porcupine Mining Region during the 19th Century Gold Rushes. (click to enlarge)

Then I did some more cool things. Using ONLY the origin coordinates, I was able to create a visualization showing the places of origin for Ideas, Objects, and People which influenced the Porcupine Gold Rush. (click to enlarge)

Later I figured out how to format the legend and titles (sorry about that).

And here’s something even more neat (that Arc CAN’T do): Change over time in an animation using the “story” feature. Note: I had to use third party software to record this, as Tableau will only let you “present” you story in the program itself.

Conclusions:

After getting over my initial annoyance about the fact that it took more than 5 seconds to learn, I liked Tableau a lot. Its easy, the learning curve is relatively short, and it can do a lot of different things (provided your data is oriented correctly). However I dislike the lack of freedom – to export stories, for example, but also to use unconventional or uncleaned data. Arc will perform the most amazing gymnastics to read spreadsheets, if you know how to use it. Tableau couldn’t even handle a csv.

Despite these shortcomings, Tableau is amazingly powerful and very customize-able. Plus, it appears to have a free version for students, so I may consider using it for my final product this summer.

 

Mica Jorgenson is a PhD candidate in the Department of History at McMaster University. She studies the environmental history of gold rushes around the globe. You can find her on twitter: @mica_amy

Posted in Blog

Leave a Reply

Your email address will not be published. Required fields are marked *

*