How to create multiple datasets for each zip code and get text clustering output

[VISHALKAPASI]

Hello Everyone..

I am using SAS Eminer 12.1 and using text parsing, filtering and finally getting results of text clustering. I am happy with unstructued data of Address, nicely classified in appropriate clusters.

Now I want this intelligent clustering within each zip code. Typically in 1 city we have more than 50 zip codes (and in addition multiple cities within country!)

The sample data lines are (For Example)

Address                                              Customer_id      Zipcode

infotech park, andheri                        1                            400701

nearby nandkamal garden, juhu        2                             400701

people colony, sion                           500                          400701

industrial tower, turbhe                     501                          400702

government quarters, belapur          502                          400702

saint international school, vashi      1000                         400702

and so on

50 zip codes (having more than 500 customers within each zipcode)

So is it possible to create a loop where SAS will create multiple dataset based on say 50 zipcodes, 50 data sets will get created or when  text clustering happens, it will create clustering grouped by zip code. So for each zip code we will have typically 20-25 clusters which we get and in final output we will have 20*50 ie. 1000 clusters, grouped by zip code

This will be really helpful. Hope I am able to explain

Regards

Vishal Kapasi

[Rick_SAS]

Although I am not familiar with the software you are using, the "SAS Way" to handle this is to use BY-group processing to analyze each ZIP code. Check your documentation to see if it has an example that demonstrates the "BY statement", "BY processing", or group processing."