There is an veritable cornucopia of troupes represented in the Ringling Circus Poster Collection. Ringling and Barnum were the given. Maybe you’ve heard of Sells and Floto. But then I found Carl Sagenbeck’s troupe and the Zallee Brothers’ and on and on. Troupe co-appearances on posters signaled an especially significant part of America circus history, but keeping names and relations straight proves tricky since collaborations were almost as frequent as mergers and takeovers.

As a starting point to understanding the overall network, data on troupes’ co-appearances in posters from the archive was uploaded to Gephi for a clearer visualization of the network.

White troupe-nodes in the network represent troupes mentioned in only one poster of the archive. The size of each node is relative to the number of times each troupe was mentioned in an archive poster. The closer each troupe-node is to another the stronger their relation is, according to their co-appearance on posters from the archive. Only troupes with 20+ mentions in the archive are labeled here.

Though these relations are often collaborative, they are even more often takeovers. For instance, the Great Wallace Show joined the Carl Hagenbeck Circus and eventually bought Hagenbeck out. Then, the American Circus Corporation bought Wallace-Hagenbeck along with several other travelling circus troupes – including Sells & Floto. Meanwhile, the Ringling Brothers merged with Barnum & Bailey who eventually bought the entirety of the American Circus Corporation and, in doing so, became the presque-monopoly of American Circus.

DATA

This site was made with data collected from The Ringling – a museum with a jumbo selection of circus artifacts including almost 5,000 vintage circus posters.

PROCESS

PYTHON: COLLECTING, CLEANING, ORGANIZING DATA

1. With BeautifulSoup, scrape each poster URL for its poster metadata.
2. Print to a JSON file containing the master list of individual poster dictionaries.
3. Using a regular expression, find posters’ publication year to establish a time range.
4. Using a regular expression, separate Troupe names from the larger title fields then loop through troupes creating a dictionary of distinct troupes and their weight based on appearance in the archive.
5. With a csv file where each row represents all the troupes appearing on one poster, loop through pairs and store weights based on frequency of co-appearance on posters from the archive.

GITHUB: TRACKING AND STORAGE

Source code is available on Github

GEPHI: VISUALIZING THE COMMUNITY

Once I had separated out troupe names I realized how wide-spread the circus business really was during this era. There are over 400 troupes represented in the archive – of various size –  and the larger ones collaborated with or were absorbed into larger troupes.

• The file made from step 4 under the Python section, can be used as a node file (with weight) to upload to Gephi.
• The file made from step 5 under the Python section, can be used as the corresponding edge file (also with weight) to upload to Gephi.