TIGER ADDRESS GEOCODING

TIGER record in database format.  Missing or non-critical fields empty.  This database was the original source for Google and Mapquest, and in many cases they still have the same problems.

The coordinates here are the bounding box, which may or may not be the end point coordinates.

  • If there are only two points in the line segment, the coordinates will be the same, but they might be in a different order.

ZIPL and ZIPR contain the ZIP codes on each side of the street.

The latest TIGER files combine FENAME and FETYPE into one field, FULLNAME.

TIGER Address Geocoding

The TIGER scheme uses four fields for each street segment to calculate address locations.

The street segment goes FROM (the first end coordinate pair) TO (the second end coordinates).

As you stand on the FROM end and look toward the TO end, there is a right side and a left side.  This results in a "From Address Left (FRADDL)" and a  "To Address Left (TOADDL)", and similar values on the right.  For each of these there is a house number, typically odd on one side and even on the other.

In this case, for the record above, the left side of the street goes from 101 to 199.  If we wanted 107, it should be about 7% of the way down the street, and the coordinates would be picked to be off the street centerline to the left.

In this case, the results are off by about two houses, because the last address on the left is nowhere near 199 (it is 109, but that house is only about 60% of the way down the block, as the last house on the block actually faces the side street). You will frequently find this level of error in addresses derived from the TIGER data (as in Google Maps and Mapquest), but this method allows quick computation for any address with a reasonable storage requirement.  If the computer addresses get you to within a few houses, that still represents a vast improvement over having to provide detailed driving directions to your visitors.

Geocoding

Geocoding in MICRODEM.


Last revision 11/9/2015