topic Re: Predefined sample selection in SAS Programming

Predefined sample selection

nathilanga — Mon, 08 Apr 2019 13:25:38 GMT

Good day

I have a two dataset as shown below

dataset1 customer_name and sitecode (has 50000 observation unique on customer_name)

sam 2430

nice 3401

jim 2030

holl 3540

and so on

dataset2 sitecode and required_number_of_customer (this dataset show the site and number of customers required per site and not repeating and this dataset contains 800 observation all unique on site code)

2430 60

4020 10

3540 2000

and so on

the requirement is that i should create one dataset that is going to contain number of unique customers according to the observation required per site

if the customer is selected to site 2430 should not be included on other site

also flag selected customers so they wont get selected on the next run

apologies i am new to SAS i currently dont have any code

Your help will be highly appreciated

thanks

Regards

Re: Predefined sample selection

Onizuka — Mon, 08 Apr 2019 13:33:39 GMT

I don't understand very well, what i understood is that you maybe need a merge.

Data dataset3 ;

MERGE dataset1 (in = A) dateset2 (in=B) ;

by sitecode ;

if A and B ;

run ;

Maybe something like that ?

Re: Predefined sample selection

ballardw — Mon, 08 Apr 2019 14:21:44 GMT

Some additional information seems likely to be needed.

Do you need to know the probability of selection or sample weight value for any of these selections?

What is the total number of the numbers of records to be selected per site code?

The example you provide for number of records needed averages over 350 per site. If that average is the same across all 800 site codes then that would mean that you need over 285000 records and that will not be possible from 50000 customer records.

Do you have any secondary rules you haven't mentioned such as if some sites may not be able to meet the goal without duplicating a customer which site the customer should be assigned to?

Re: Predefined sample selection

nathilanga — Tue, 09 Apr 2019 13:04:29 GMT

Good day

this is merging my goal out of lets say i have 200 site with and summed required customers for all sites 3000 and my base of customers is 10000 i should have 7000 flagged as not selected one more time

customerbase has below examples

customername sitecode

ben 2430

get 1000

sip 2430

site dataset

sitecode numberofcustomersrequiredpersite

2430 10

1000 8

all customer that should be under 2430 for an example in a base table they must be on site 2430

Re: Predefined sample selection

nathilanga — Tue, 09 Apr 2019 13:05:31 GMT

good day

one line item I didnt mention is below

all customer that should be under 2430 for an example in a base table they must be on site 2430