do loops or do until with the by statement

[#1]

Hi, 

I have a dataset with some accounts and every account may have more than one observation in the data. 

Assuming a scenarion, if one of the account has 10 observations and every observation has a different stat_code which is one of the variables along with stat_dt

What i want to do is to go through the first entry till the last entry of the duplicate accounts and then apply a condition 

on stat_code, something like if substr(stat_code, 4,1) = 'S' and every time it finds that between the first and the last observation, it outputs that entry into a seperate dataset. 

Will appreciate some direction and help!

--Tej

[#2]

Please provide sample (in machine readable format) input and expected output data, for us to work on.

[#3]

Here is a sample input. I want to go through each and every account as you see there are duplicate and search for where substr(status_code,4,1) = 'L' and where ever it finds that it outputs that record.

Account	Status Code	Date
11112	XYZS	10th Jan 2011
11112	HGJL	11th Jan 2011
11112	FTYJ	12th Jan 2011
11112	LKIL	13th Jan 2011
234322	PLKS	6th June 2016
234322	RTYZ	7th June 2016
234322	KJLL	8th June 2016
234322	IUYL	9th June 2016
234322	RTYZ	10th June 2016
234322	RTYZ	11th June 2016
234322	RTYZ	12th June 2016

Output should look like below as it output recoed whereever it finds that string

11112	HGJL	11th Jan 2011
11112	LKIL	13th Jan 2011
234322	KJLL	8th June 2016
234322	IUYL	9th June 2016

[#4]

Did you try anything? This should be really straight forward, especially after the video example.

[#5 · Accepted solution]

I don't see anything to indicate the need for something besides a simple filter. 

SAS loops through each one in the dataset by default, and you can implement BY group processing if it's needed. 

See the tutorial here in how to filter

http://support.sas.com/training/tutorial/studio/filter-tables.html

[#6]

I know i can do the where clause but I am also looking to figure out a way to do this by "by processing"

[#7]

BY processing usually means there is something special done with each level or group, or combination of levels, of the By variable(s).

You do not show anything that is dependent on a group of values.

I could add a by statement to the datastep but it wouldn't add any value. If you wanted to know something like the order number within the account value then that could use by processing. 

If there is something else other than the data shown and the rule  and desired results then there may be some value to adding by group processing.