FangSoft

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38

private static final char[] HEX_CHARS = "0123456789abcdef".toCharArray();
 
public static String generateHash(File file, String algorithm) throws NoSuchAlgorithmException, IOException {
	byte[] dataBytes = getFileContents(file);
	MessageDigest messageDigest = MessageDigest.getInstance(algorithm);
	messageDigest.update(dataBytes);
	byte[] digestBytes = messageDigest.digest();
 
	// this chunk of code was taken from http://echochamber.me/viewtopic.php?f=11&t=16666&p=553936#p459685
	char[] hash = new char[2 * digestBytes.length];
	for (int i = 0; i < digestBytes.length; ++i) {
		hash[2 * i] = HEX_CHARS[(digestBytes[i] & 0xF0) >>> 4];
		hash[2 * i + 1] = HEX_CHARS[digestBytes[i] & 0x0F];
	}
 
	return new String(hash);
}
 
private static byte[] getFileContents(File file) throws IOException {
	FileInputStream fileInputStream = null;
 
	try {
		fileInputStream = new FileInputStream(file);
		byte fileBytes[] = new byte[fileInputStream.available()];
		fileInputStream.read(fileBytes);
		return fileBytes;
	} catch (IOException e) {
		throw e;
	} finally {
		if (fileInputStream != null)
			try {
				fileInputStream.close();
			} catch (IOException e) {
			} finally {
				fileInputStream = null;
			}
	}
}

I edited the KML ZCTA population map that I posted on 2/22/09. For my uses, it’s important that I be able to calculate distances between ZCTAs. To calculate these distances, I use the center of each ZCTA. To avoid complex centroid calculations, and since the Census Bureau’s ZCTA boundary data includes the centers so that this calculation isn’t necessary, I figured I’d include them.

I did this by changing the KML geometry from a single LinearRing to a MultiGeometry of LinearRings and Points.  Let me know if this is useful to you!

Also, if you’re curious, the distance formula for computing the distance (in miles) between two latitude/longitude points (in degrees) is fairly simple. Courtesy, http://www.meridianworlddata.com/Distance-Calculation.asp, here it is:

Download: http://www.fangsoft.net/public/populationZCTA.kmz (20.8mb)

Here’s what this new file looks like in Google Earth:

My research thus far in grad school has been concentrated around using computer science and GIS tools/techniques to help solve medical problems.  For one of the problems I’m working on, I need access to ZCTA boundaries and population counts for each ZCTA.  For those of you that don’t know, ZCTA stands for ZIP-code tabulation area; basically, it’s what the Census Bureau uses to split the US population up geographically.  ZCTAs typically correspond to ZIP codes, but unfortunately, that’s doesn’t always happen (there is a ZCTA 506HH, for example).  Some ZCTAs have populations of 0.  However, as far as I can tell, ZCTAs are the best thing we have for geographically representing the US population.

Anyway, what I was doing was using the shapefiles publicly available at http://www.census.gov/geo/www/cob/z52000.html to get the ZCTA boundary information.  I also used the 2000 census population data available from http://factfinder.census.gov/servlet/DownloadDatasetServlet?_lang=en. This is all fine and dandy, but reading shapefiles can be kind of a pain.  GeoTools works quite well, but it’s Java-only.  I’m in the process of converting a Java-based application into a PHP web-based application for some of my research.  PHP’s shapefile readers, at least from what I’ve heard, can be kind of hit or miss.  Also, having to get this data from multiple sources starts to make code look kinda nasty.  Ideally, since population information goes with ZCTAs, there should be a 1-stop shop that gives me both the boundary information and population counts for each ZCTA in the US.  And if it’s not too much to ask for, I’d like a human-readable text format.  So what do I do?

A KML map seemed like the most obvious answer.  It’s XML, so it’s easily readable, and it’s highly extensible and powerful.  And since Google’s name is behind it, it’s very widely supported.  What’s not to like?  Surely, a KML file like this exists, right?  Well, after quite a bit of Googling, I came to the conclusion that one didn’t exist.  So, I decided to make my own.  I wrote a Java program to read in the population information and ASCII coordinates of each ZCTA listed at the bottom of the Census Bureau’s page I linked to in the above paragraph and spit out a KML map.

Because this is so useful to me (and really, anyone doing anything that requires ZCTA-level population counts), I figured I’d upload it here and let anyone that wants it grab it.  This took me a decent chunk of time to get done, so maybe I can save someone else some time by sharing this.  If you use this, please let me know by leaving a comment.  Also, if you’re interested in the Java code I used to create the KML file, feel free to ask.

Download: http://www.fangsoft.net/public/populationZCTA.kmz (size: 20.1mb)

Note: This file has a KMZ extension (not a KML extension).  A KMZ file is simply a zipped-up KML file.  If you wish to access the underlying KML markup, simply rename the file to populationZCTA.zip, and unzip it like you would a normal ZIP file.  Google Earth can read the KMZ file without unzipping.

Here’s what it looks like in Google Earth: