Hacker News Reader

Comment by spodek | Hacker News Reader

Here is the actual map, which you can zoom into, scroll, pan, explore your neighborhood, etc. -- http://demographics.coopercenter.org/DotMap/index.html

EDIT: Exploring, I see the island I live on -- Manhattan -- has the four main dot colors well represented, though mainly in blocks. The main mixing is between whites and Asians and between blacks and Hispanics. Notably parts of the Upper West Side, East Village/Lower East Side, and Midtown west show significantly more mixing, at least by eye -- it would be interesting to quantify the data.

Some parts of Queens and Brooklyn show mixing of all four dot colors within the same city block.

Still, on the whole most dot colors remain in blocks with less mixing on the edges. Some dot colors rarely mix with other dot colors.

EDIT TWO: How they made the map is here -- http://www.coopercenter.org/demographics/Racial-Dot-Map -- which describes the data sources and links to the code on Github.

View on HN

The actual map is an amazing piece of work, but let's give credit where credit is due!

Bill Rankin posted maps in this exact visual style back in 2009. You can see them (and many other amazing maps he's done) at his blog -- http://www.radicalcartography.net/index.html?chicagodots

Eric Fischer picked up Bill's project and ran with it, composing maps for most of the major US cities with census 2010 data. You can see them all on his Flickr page -- http://www.flickr.com/photos/walkingsf/sets/7215762481267496....

I love great maps, but it's a little disheartening that these precursor projects aren't mentioned anywhere in the article/comments.

Thank you. I saw this article and had a feeling of "Hmm, where have I seen this before? The URL says 2013, but this is definitely old..." I think it must have Eric's flickr.

(Link somehow got truncated; it should be http://www.flickr.com/photos/walkingsf/sets/7215762481267496... )

The website mentions extracting "308,745,538 observations" (21 GB in size) and transforming it to some custom CSV files, for processing via Java. I don't see any of the Java code on GitHub.

...he mentions use of a summary file, "SF1", and there's some kind of PDF here:

--> http://www.census.gov/prod/www/decennial.html#y2010tecdoc

--> http://www.census.gov/prod/cen2010/doc/sf1.pdf

It's more or less, a user manual for roughly 300 other tables, but where are they? Is the actual census data itself available freely, in detail?

I think this is it:

--> http://www2.census.gov/census_2010/04-Summary_File_1/

New York, as an example:

--> http://www2.census.gov/census_2010/04-Summary_File_1/New_Yor...

Is this the original, actual source data?

I have some admittedly old Python code to work with Census data here: https://github.com/peterlandry/django-census

The small talk I gave introducing python geeks to census data is here: http://files.meetup.com/3119472/census%20data%20for%20python...

I'm glad the code is available. I'd like to see this visualization applied to previous census data, so that we can visually compare two points in time (or more). So, say:

* the 2000 census data

* the 1990 census data

* the 1980 census data

... and so on. If this can be visualized, and then maybe animated, that would provide some insightful results.

Paging http://en.wikipedia.org/wiki/Hans_Rosling, paging Hans Rosling. Animated visualization of statistical data required, please respond ...

Very fascinating. I wish you could link to specific map views. I'd love to send direct links to different areas.

Notice how there is often a latino or Asian neighborhood between the white and african american neighborhoods.

I'd like to see figures on racial division in the labor market. When I lived in NYC I couldn't help noticing how I would rarely see white people working in stores, on the streets or driving cabs (and if they were white they would be immigrants). Diners were mostly staffed with Greek people, except the guys who poured water and cleaned up after you - they would be latino. White New Yorkers mainly work in offices. Waiters in restaurants or bars are often white as well.

Interestingly, in the D.C. area, you'll see large areas to the West (Virginia) with Asian/Hispanic mixes, large areas to the East (Maryland) of all racially Black areas, and North (Maryland) non-mixed densely packed blocks nearby for all races.

In the Western areas, there are very large Korean and Latin mixed population areas, so much so that most Korean restaurants are owned by Koreans, but all the cooks and staff are Latin American and the staff speaks Spanish and a little Korean, but no English and the owners are the reverse. In some ways, it's a little like K-Town in LA, known as "Korea town" but with large Hispanic minority groups mixed in. Also the home of the "Bulgogi Taco".

It's a very unexpected integration that somehow works pretty well.

The Map is really cool. I wish you could zoom in more. It's hard to find my actual block on manhattan.

Neighborhoods doesn't tell the whole story in New York because we have a subway. We are integrated at the breating space level on a daily basis.

Riding the subway means that as a white New Yorker, you will twice a day share a room with a black person. Whether you interact with them is another story - I'd venture that most people never say a word to anyone in the subway. To understand (likelihood of) interaction, neighborhood maps are one piece of the story, as would be (as mentioned elsewhere in the thread) workplace data.

I hardly talk to my neighbors too. I do most talking with my friends and coworkers.

But you are also likely to live in the same neighborhood as your friends. So while living in a racially diverse neighborhood may not cause you to interact with people of other other races, it may be indicative if your group of friends is racially diverse.

Reply on news.ycombinator.com