This project is focusing on mapping public schools in New York City using Tableau. The original idea was to look at car accidents involving pedestrians and then identify those that occurred near public schools. However, that proved to be extremely difficult to accomplish using only Tableau, so the final decision was to understand how mapping works on this software doing a simpler task.
The inspiration for these graphs comes from a similar project that I have done in my Intro to GIS class. This class is all about using shape files and other geo-location data to make maps on QGIS, a software that is used for mapping and analyzing geospatial datasets. A project that I completed for that class was about mapping MBTA stations as dots, as well as use different color to show the polygon of each borough on the map. Another inspiration is from the NYC Department of Education website. They have an interactive map that allows people to look for schools of their choice near the certain address and such, and even though it’s hard to do a complicated one like that, it would be interesting to display of of the information they have on this map. So for this project, the goal is to map out all the public schools in NYC, color them by their categories, as well as categorize and color the each school district polygon by the number of public schools it has.
The datasets used for this project were retrieved from NYC Open Data. One of them is the School Point Locations, which contains all the coordinates for all public schools; the other one is School Districts, which is the shape file of the school districts. Both files are downloaded as spatial files so that they can be joined and analyzed in Tableau. After downloading these files, the first task was to combine these two spatial files using the admin district and the school district fields. This process took a while, since the two spatial fields that needed to be combined were in different formats. Even though they were both whole numbers, Tableau categorized one of them as discrete and the other one as continuous. This mean that they won’t relate with each other. With some help, this issue was resolved by making both of them as strings so that which kind of number they were doesn’t matter anymore.
After combining these two shape files, it was time to map the school district as well as the school locations. The first step was to use the generated longitude (columns) and latitude (rows) to set the map to New York area, then use the geometry field in the district map to create the boundaries for school districts by adding it to details. In order to map the locations as dots on the district, another map layer was added using the school point locations file. For this layer, geometry field was added to detail so each school was shown as a dot on the map, then the school type filed was added to color so each dot could represent a different school type. Borough was then added to filter to get rid of all the null values, making sure that all schools were in the boundaries. A legend was then generated, containing 8 different types of schools, and readers can look at each individual school types by clicking on to the legend.
This map shows that elementary school is the main type of public schools in all school districts and there are not a lot of early childhood public schools. Another interesting fact is that there is one school that is ungraded in midtown Manhattan.
Another map was then created to school districts and the number of schools each of them contains. This map was created using the same spacial files with the same method of combing them. After putting in the longitude and latitude again, both school district number and the geometry of school districts were added to the map, so that polygons of school districts were created. This means that when audiences move the mouse on to the map, each school district would pop-up, showing its number at the same time. After that, the total count of schools in the NYC area were added to color, and now each school district polygon were gradually colored based on the count of public school it has, with lightest blue being the lowest and deepest blue being the highest.
School district 2 in Lower Manhattan with 117 public schools definitely has the highest number, and the lowest school district 32 with 29 public schools. Staten Island itself is a whole entire school district, which is unusual.
These two maps really shows that different mapping methods works for different purposes. It would be very chaotic if dots were added on to map layer that also has color indicating how many school a polygons has, since there will be a lot of colors, and it would also be very weird if you are trying to color the school district based on the type of school it has. It is a different yet interesting experience using spacial files to map in Tableau, since I use QGIS most of the time. Tableau treats spacial file in the same way it treats any other datasets, and it has a very uniform language across the board no matter what kind of data you are trying to visualize. It could be an advantage since once you learned its language you can navigate all kinds of data in a similar way; however, it could also be a draw back since some mapping methods or datasets do require a more concentrated, professional way of exploring and analyzing.
It would be interesting to know to how to accomplish more complicated spacial data mapping in Tableau. For example, going back to the initial question of “how many pedestrians were involved in a car accident and injured near different public school.” To map this in Tableau could take many steps.