is this correct??

[Markanti]

1. Use the code in the unfinished code section to create the data set call “two” from “one”, which you created in problem 1.  Add the following to “two”:
   1. Create a new text variable called “yrob2” using the last 2 digits of the variable “yrob”.  For example if year = 1980 we would want 80 for this new variable. 
   2. Next, create a new text variable called “last4” from SSN using the last 4 digits of the SSN.
   3. Now, create a new text variable called “ID” by concatenating “clinic”, “mthob”, “yrob2” & “last4” (in this order). 
   4. Remove any spaces or blanks from this new variable called “ID”. 
   5. Remove all variables from this dataset except for name & ID. Hint: Think about the most efficient way to do this.

finished Code

*Lab 8 – Problem 2;

title "Building an ID number from data";

data lab.ssn_number;
set lab.one;
ssn2 = compress(ssn, '-');
ssn3 = input(ssn2, ??11.);
drop ssn2;
run;

proc print data=lab.ssn_number;
run;

[Patrick]

The code you shared doesn't have much to do with solving problems 1 to 5. 

To support homework forum members normally want to see some of your own effort and then help with the bits where you get really stuck after you first tried yourself. 

[ballardw]

Hint: When you have text values with a delimiter such as - that usually appears in SSN then SCAN is useful tool to get the nth element of the value

Hint2: NEVER input parts of character values to numeric if the instruction is to Concatenate. You will lose leading zeroes and add complication to the code for concatenating.

Personally I consider #3 to be dangerous in terms of actual use. Discuss this with your instructor.

Suppose you have two values before removing blanks that look like this (because you created a numeric value from SSN of 0078 that becomes 78 for example)

' 1 2 12 5678'
'12 12 56 78'

If you remove blanks you get

'12125678'
'12125678'

Unless your codes for things like Clinic and your month always uses 2 digits (and years REALLY should use 4 digits, look up Y2K) you have a possibility of creating duplicate values that should mean different things. Of course I do not have any of your data to see how likely this is but is something to be aware of.