Earlier this week we released the Data Observatory. We built the Data Observatory because we wanted easier to find and easier to use geospatial data. More than that, we wanted to put in place the foundation for some things we have in store for the future (more on that in my next blog post). We are all a bunch of map and data nerds, so this is a project that really came from our hearts. While the Data Observatory is far from complete, we built it so it could expand rapidly into new areas of data, the first release is enough to have us tearing up a bit. We think it has the power to change the way you analyze location data and potentially the way you think about maps entirely.
Here are 5 ways we think it’s going to change your relationship with CartoDB:
1. CartoDB will become the first place your look for the staples of location data.
The Data Observatory’s list of available data grows bigger every week. Our focus is on the high-value and/or hard-to-use datasets that people and businesses ask us for all the time. For people who need or just love location data, the Data Observatory might become a bit of an addiction. Be careful!
Example
Grab U.S. Census Block Groups for your home town quickly by simply running this function and clicking “create dataset from query” in the CartoDB Editor.
sql
SELECT *
FROM OBS_GetBoundariesByGeometry(
ST_Buffer(CDB_LatLng(40.689, -73.944),0.1),'us.census.tiger.block_group')
2. Soon, you’ll be wondering if you normalize data too much.
It is now super simple to normalize your data in CartoDB with some of the world’s most trusted sources of denominators. So why not see what your data looks like per household? Or, per person? Or, per income($)? It’s so easy you might as well take a look. :)
Example
Add the U.S. population, per square meter, as a new column of any point dataset in your CartoDB account. Simply add a new column called, “population”, and run the following statement,
sql
UPDATE SET
population = OBS_GetUSCensusMeasure(the_geom, 'Total Population')
3. You will start making maps of people just because.
This is one that we know to be true. We know it because we’ve been living it for the past two months. Our team has had the pleasure of early access to the Data Observatory. The result of that access? Maps. Maps for curiousity. Maps for discovery. Maps just because. It’s awesome and you’re going to love it. There are so many interesting measures in the Data Observatory, it’s worth just playing around.
Example
Take for example, the number of people who walk to work as their primary mode of transportation. You can add the count of these people to any of your U.S. based polygons simply by adding a new column, “walkers”, to your table and then running,
sql
UPDATE SET
walkers = OBS_GetUSCensusMeasure(the_geom, 'Walked to Work')
4. Basemaps will become more and more optional.
One thing that I’ve noticed a lot since we started getting more and more data into the Data Observatory is that our Senior Cartographer started having fewer and fewer basemaps in her visualizations. And even those that do use a basemap are often actually just using boundary layers from the Data Observatory to replace the fully detailed basemaps we all normally use.
Example
Use a new boundary dataset as a blank canvas instead of a basemap. For example, grab the U.S. Congressional Districts and add them to a new blank table. First, create a new empty dataset in CartoDB. Add a new string based column called, ‘geoid’. Rename your new table, ‘us_congressional_districts’ and just run,
sql
INSERT INTO us_congressional_districts(the_geom, geoid)
SELECT the_geom, geom_ref
FROM OBS_GetBoundariesByGeometry(
ST_MakeEnvelope(-179.5, 13.4, -42.4, 74.4, 4326),'us.census.tiger.congressional_district')
5. You’ll find many new stories in old data.
In our mapping rampage we found ourselves thinking this quite a lot. The Data Observatory has actually backed some of our recent blog posts, including the L Train Analysis, Airbnb Impact Mapping and Data Mountains. What we now know is that we have only scraped the surface of the stories out there.
Let’s go dive into data!
Happy data mapping!