University:
Boston University
Course:
QST SM 101 | Introduction to Management
Academic year:

2024
Views:

325

Pages:

24
Author:

Tara Guerrero

Shorten your SAS Code with Character Functions Manipulating Character Strings Boston University Slideshow Title Goes Here Concatenate character strings without removing leading or trailing blanks Old = X1 || X2 || X3 || X4; New = CAT(OF X1 – X4); X1 This X2 is X3 a X4 Old sentence This is a sentence New This is a sentence 2 Manipulating Character Strings Boston University Slideshow Title Goes Here Concatenate character strings and remove trailing blanks Old = TRIM(X1) || TRIM(X2) || TRIM(X3) || TRIM(X4); New = CATT(OF X1 – X4); X1 This X2 is X3 a X4 Old sentence This is asentence New This is asentence 3 Manipulating Character Strings Boston University Slideshow Title Goes Here Concatenate character strings and remove leading and trailing blanks Old = TRIM(LEFT(X1)) || TRIM(LEFT(X2)) || TRIM(LEFT(X3)) || TRIM(LEFT(X4)); New = CATS(OF X1 – X4); X1 This X2 is X3 a X4 Old sentence Thisisasentence New Thisisasentence 4 Manipulating Character Strings Boston University Slideshow Title Goes Here Concatenate character strings, remove leading and trailing blanks, and insert separator Old = TRIM(LEFT(X1)) || “ ” || TRIM(LEFT(X2)) || “ ” || TRIM(LEFT(X3)) || “ ” || TRIM(LEFT(X4)); New = CATX(“ ”, X1 – X4); X1 This X2 is X3 a X4 Old sentence This is a sentence New This is a sentence 5 Manipulating Character Strings Boston University Slideshow Title Goes Here Note: default length for variables created using CAT functions is different from the length that is obtained when using the concatenation operator ( || ) data names; input last : $25. first : $15. @@; full_old = trim(last) || ', ' || trim(left(first)); full_new = catx(', ', last, first); datalines; Sisson Emily Palmisano Joe Coleman Sharon ; run; 6 Manipulating Character Strings Boston University Slideshow Title Goes Here Results: Obs last first full_old full_new 1 Sisson Emily Sisson, Emily Sisson, Emily 2 Palmisano Joe Palmisano, Joe Palmisano, Joe 3 Coleman Sharon Coleman, Sharon Coleman, Sharon Alphabetic List of Variables and Attributes # Variable Type Len 2 first Char 15 4 full_new Char 200 3 full_old Char 42 1 last Char 25 7 Search for Variable Values Boston University Slideshow Title Goes Here The next few examples will consider a dataset housing answers to a 10-question survey: id a1 a2 a3 a4 a5 a6 a7 a8 a9 a10 1234 Y Y Y Y Y Y Y Y N N 2345 N N N N N N N N N N 3456 N N N N N N N N N Y 4567 N N N N Y N N N N N 5678 Y N N Y N N Y N N Y 8 Search for Variable Values Boston University Slideshow Title Goes Here Using the survey dataset, create a new dataset containing observations where at least one answer was ‘Y’. Method 1: Brute Force: data yes; set survey; if a1 = 'Y' or a2 = 'Y' or a3 = 'Y' or a4 = 'Y' or a5 = 'Y' or a6 = 'Y' or a7 = 'Y' or a8 = 'Y' or a9 = 'Y' or a10 = 'Y'; run; 9 Search for Variable Values Boston University Slideshow Title Goes Here Using the survey dataset, create a new dataset containing observations where at least one answer was ‘Y’. Method 2: Conventional Array/Do Loop: data yes; set survey; array a(10); do i = 1 to 10 until (a(i) = 'Y'); end; if i < 11; run; 10 Search for Variable Values Boston University Slideshow Title Goes Here CAT and FIND Functions: data yes; set survey; if find(cat(of a1-a10), 'Y') > 0; run; id a1 a2 a3 a4 a5 a6 a7 a8 a9 a10 cat(of a1-a10) 1234 Y Y Y Y Y Y Y Y N N YYYYYYYYNN 2345 N N N N N N N N N N NNNNNNNNNN 3456 N N N N N N N N N Y NNNNNNNNNY 4567 N N N N Y N N N N N NNNNYNNNNN 5678 Y N N Y N N Y N N Y YNNYNNYNNY 11 Search for Variable Values Boston University Slideshow Title Goes Here Using the survey dataset, create a new dataset containing observations where at least TWO answers were ‘Y’. Conventional Array/Do Loop: data yes; set survey; array a(10); do i = 1 to 10; found_Y = sum(found_Y, (a(i) = 'Y')); end; if found_Y >= 2; drop i found_Y; run; 12 Search for Variable Values Boston University Slideshow Title Goes Here CAT and COUNTC Functions: data yes; set survey; if countc(cat(of a1-a10), 'Y') >= 2; run; 13 Search for Variable Values Boston University Slideshow Title Goes Here Character functions can also come in handy for numeric data! Consider a dataset housing answers to a 10-question survey: id a1 a2 a3 a4 a5 a6 a7 a8 a9 a10 1234 5 5 3 3 2 1 1 1 1 1 2345 0 0 0 0 0 0 0 5 5 5 3456 1 1 2 2 4 0 0 0 0 0 4567 5 5 5 5 5 5 5 5 5 5 5678 1 0 1 0 1 0 5 5 5 3 Task: Using the numeric survey dataset, create a new dataset containing observations where at least one answer was 5. 14 Search for Variable Values Boston University Slideshow Title Goes Here CAT and FIND Functions: data five; set surveynum; if find(cat(of a:), ‘5’); run; id a1 a2 a3 a4 a5 a6 a7 a8 a9 a10 cat(of a:) 1234 5 5 3 3 2 1 1 1 1 1 5533211111 2345 0 0 0 0 0 0 0 5 5 5 0000000555 3456 1 1 2 2 4 0 0 0 0 0 1122400000 4567 5 5 5 5 5 5 5 5 5 5 5555555555 5678 1 0 1 0 1 0 5 5 5 3 1010105553 15 Practical Application: Diagnosis Codes Boston University Slideshow Title Goes Here Each hospital admission can have up to five three-character diagnosis codes. Find patients with Diabetes (any DX code of 250). Patient DX1 DX2 DX3 DX4 001 025 022 002 682 401 244 493 003 592 401 493 004 428 493 780 V43 005 250 006 414 V45 401 250 DX5 250 16 Practical Application: Diagnosis Codes Boston University Slideshow Title Goes Here Why does find(cat(of DX1-DX5), ‘250’) not work in this situation? Patient DX1 DX2 DX3 001 025 022 002 682 401 244 003 592 401 493 004 428 493 780 005 250 006 414 DX4 DX5 cat(of DX1-DX5) 025022 493 682401244493 592401493 V43 250 428493780V43250 250 V45 401 250 414V45401250 17 Practical Application: Diagnosis Codes Boston University Slideshow Title Goes Here Why does find(cat(of DX1-DX5), ‘250’) not work in this situation? Patient DX1 DX2 DX3 001 025 022 002 682 401 244 003 592 401 493 004 428 493 780 005 250 006 414 DX4 DX5 cat(of DX1-DX5) 025022 493 682401244493 592401493 V43 250 428493780V43250 250 V45 401 250 414V45401250 18 Practical Application: Diagnosis Codes Boston University Slideshow Title Goes Here Solution: Use CATX to insert a separator! find(catx(‘*’,of DX1 – DX5), ‘250’) Patient DX1 DX2 DX3 001 025 022 002 682 401 244 003 592 401 493 004 428 493 780 005 250 006 414 DX4 DX5 catx(‘*’, of DX1-DX5) 025*022 493 682*401*244*493 592*401*493 V43 250 428*493*780*V43*250 250 V45 401 250 414*V45*401*250 19 Practical Application: Diagnosis Codes Boston University Slideshow Title Goes Here Of course, ICD codes are usually a little more complicated than all this. The first 3 digits indicate a disease, and 0-2 digits may follow for a further-refined diagnosis… Patient DX1 DX2 DX3 DX4 DX5 001 25001 V180 002 5680 78039 3250 49390 53081 003 4270 4111 4019 36250 004 78659 25000 49320 005 49392 4660 2449 006 34839 2765 40493 4280 20 Practical Application: Diagnosis Codes Boston University Slideshow Title Goes Here Why do these statements not work to find Asthma (starting with 493) or Diabetes (starting with 250)? diabetes = (find(catx('*',of dx1-dx5),'*250') gt 0); asthma = (find(catx('*',of dx1-dx5),'*493') gt 0); Patient DX1 DX2 DX3 DX4 DX5 catx('*',of dx1-dx5) 001 25001 V180 002 5680 78039 3250 49390 003 4270 4111 4019 36250 004 78659 25000 49320 78659*25000*49320 005 49392 4660 2449 49392*4660*2449 006 34839 2765 40493 25001*V180 4280 53081 5680*78039*3250*493390*53081 4270*4111*4019*36250 34839*2765*40493*4280 21 Practical Application: Diagnosis Codes Boston University Slideshow Title Goes Here Why do these statements not work to find Asthma (starting with 493) or Diabetes (starting with 250)? diabetes = (find(catx('*',of dx1-dx5),'*250') gt 0); asthma = (find(catx('*',of dx1-dx5),'*493') gt 0); Patient DX1 DX2 DX3 DX4 DX5 catx('*',of dx1-dx5) 001 25001 V180 002 5680 78039 3250 49390 003 4270 4111 4019 36250 004 78659 25000 49320 78659*25000*49320 005 49392 4660 2449 49392*4660*2449 006 34839 2765 40493 25001*V180 4280 53081 5680*78039*3250*493390*53081 4270*4111*4019*36250 34839*2765*40493*4280 22 Practical Application: Diagnosis Codes Boston University Slideshow Title Goes Here Solution: diabetes = (find(catx('*', '*', of dx1-dx5),'*250') gt 0); asthma = (find(catx('*', '*', of dx1-dx5),'*493') gt 0); Patient DX1 DX2 DX3 DX4 DX5 catx('*',’*’,of dx1-dx5) 001 25001 V180 002 5680 78039 3250 49390 003 4270 4111 4019 36250 004 78659 25000 49320 *78659*25000*49320 005 49392 4660 2449 *49392*4660*2449 006 34839 2765 40493 *25001*V180 4280 53081 *5680*78039*3250*493390*53081 *4270*4111*4019*36250 *34839*2765*40493*4280 23 Character Functions Used: Summary Boston University Slideshow Title Goes Here  CAT Operators (CAT, CATT, CATS, CATX)  Useful in evaluating multiple character strings simultaneously  FIND  Allows you to identify the first position in a string that contains the specified search term  COUNTC  Counts number of characters in a string that contain the specified search character  Others:  http://support.sas.com/publishing/pubcat/chaps/59343.pdf 24