Sure, we could (and did!) make choropleths of the data at various geographic levels, but a choropleth at the block level ignores the fact that we know the specific dwellings in which individuals lived.Īnd choropleths are effective when looking at one value within a dataset (such as the percentage of people born abroad), but once you try to look at more than two values in the dataset at once (say, the percentage of Irish-born people AND the percentage of Russian-born people AND the percentage of Caribbean-born people) choropleths quickly become difficult to read and unfit for the task. The scale and specificity of the data introduced a number of interesting challenges, and we found that many of the more obvious ways of mapping the data were unsatisfying considering how detailed the data was. ![]() We at Stamen were among the lucky first to visualize this rich data. An example page from the 1880 US Census via Wikimedia Commons. Indeed, the Center for Spatial Research spent a few years laboriously cleaning this data (for the 1850, 1880, and 1910 censuses) and associating each individual in the data with a dwelling and associating that dwelling with a geographic point. This data is available historically because the US Census Bureau releases data at the individual level after 72 years, but that data isn’t necessarily available digitally or associated with building points. Many of us who make maps in the US have dealt with census data that is aggregated to the census block or tract, and it’s a rare treat to be able to include data that is more granular than that. You can also hear us talk to Dan Miller from the Center for Spatial Research on our podcast episode about the project. In this post we’ll talk about other methods we attempted along the way and share some of the technical details about how we created this map. We ended up visualizing the data with a variation on a dot density map (seen above) that has some unique effects when you zoom into the street level, as we’ll show below. ![]() Population by occupation in 1880 via Mapping Historical New York. Part of what is special about this data is that in addition to being aggregated at various geographical units (such as city blocks and wards), the data is also available at the building and individual levels. ![]() The Center came to the table with a large and unique dataset of historical census data for the areas that are now Manhattan and Brooklyn dating back to 1850. In 2021 Stamen had the pleasure of working on Mapping Historical New York with Columbia University’s Center for Spatial Research.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |