Anyone tried doing any geocoding type work in SAS without using the GIS product?
By geocoding I mean working with address data and assigning things to geographic locations?
Primarily my interest at the moment is in trying to find a way to find the nearest point to another point, for instance, locating the nearest police station to a house which I have the latitude and longitude for.
I've been able to find the algorithms for calculating distances between points, however the issue I now face is how do I do this for large datasets (ie, I have 9 million addresses, and 10's of thousands of police stations scattered across Australia, if I calculate the distance from every address to every police station and then take the smallest one, it's going to take me all week to process the data.
Has anyone got any ideas on how to handle this?
I did have an idea of dividing the country up into a grid and building an array of the centroids of each square and using this in some sort of iterative process to identify the closest grid square to reduce the number of joins required.
Has anyone tried anything like this? Has anyone used SAS/GIS and have any insight on what it's like?
Chris
Business Analyst, Code Monkey, Data Wrangler.
SAS Guru.
By geocoding I mean working with address data and assigning things to geographic locations?
Primarily my interest at the moment is in trying to find a way to find the nearest point to another point, for instance, locating the nearest police station to a house which I have the latitude and longitude for.
I've been able to find the algorithms for calculating distances between points, however the issue I now face is how do I do this for large datasets (ie, I have 9 million addresses, and 10's of thousands of police stations scattered across Australia, if I calculate the distance from every address to every police station and then take the smallest one, it's going to take me all week to process the data.
Has anyone got any ideas on how to handle this?
I did have an idea of dividing the country up into a grid and building an array of the centroids of each square and using this in some sort of iterative process to identify the closest grid square to reduce the number of joins required.
Has anyone tried anything like this? Has anyone used SAS/GIS and have any insight on what it's like?
Chris
Business Analyst, Code Monkey, Data Wrangler.
SAS Guru.