SAS® Cheat Sheet
SAS Language
ATTRIB varn <LENGTH=’varn-length’> <LABEL=’varn-label’>
<FORMAT=varn-format.> <INFORMAT=varn-informat.>;
Associates a format, informat, label, and/or length with one
or more variables
CARDS or CARDS4 | DATALINES or DATALINES4
Indicates that data lines follow (suffix of 4 if data has ‘;’s).
DATA <dsetn <(dset-optionsn)>>;
Begins a DATA step and provides names for any output
SAS data sets. See Data Set Options for options that are
available in the DATA statement.
DO index-var=start_value TO end_value <BY step>;
DO UNTIL (expression);
DO WHILE (expression);
Groups a set of statements as a single unit. Note that
UNTIL conditions are evaluated at the end of the loop and
thus execute at least once.
FILE filename <options>;
Specifies the current output file for PUT statements.
Options include:
MOD output is appended to an existing file.
OLD output overwrites an existing file.
IF expression THEN statement; .... <ELSE> statement;
SAS evaluates the expression in an IF statement to
produce a result that is either non-zero, zero, or missing. If
result >0 then TRUE, else FALSE.
INFILE filename <options>;
Specifies an external file to read with an INPUT statement.
Options include:
DELIMITER|DLM= delimiters
Specifies a delimiter for list input.
LENGTH= variable
Names a variable that SAS sets to the length of the
current input line.
INPUT var<=> <$> startcol <-endcol> <.dec> <@ | @@>;
INPUT <pointer-control> variable informat. <@ | @@>;
INPUT <pointer-control> variable <$> <&> <@ | @@>;
Input records from the current input file, placing the values
into SAS variables.
MERGE ds1 <(doptions)> <... dsn<(doptions)>> <END=var>;
Joins observations from two or more SAS data sets into
single observations.
OUTPUT<data-set-name(s)>;
Writes the current observation to a SAS data set.
PUT var<=> <$> startcol <-endcol> <.dec> <@ | @@>;
PUT <pointer-control> <”text”|variable format.> <@ | @@>;
Writes variable values and/or text to the output line.
RETAIN variablen <initial-valuen>;
Causes a variable to retain its value from one iteration of
the data step to the next.
SET <data-set(s) <(data-set-options(s)>> <POINT=varname>
<NOBS=varname> <END=varname>;
Reads observations from one or more data sets.
Sum: variable+expression MISSING(argument) Indicates whether the argument
Adds the result of an expression to an accumulator var. contains a missing value
TITLE <n> <”text”>; MOD(argument1, argument2) Returns the remainder
Specifies title lines for SAS output. n specifies the relative MONTH(date) Returns the month from a SAS date value
line number with n being between 1 and 10. RANK(x) Returns the position of a character x in the ASCII
WHERE where-expression; or EBCDIC collating sequence
Selects observations from SAS data sets that meet a REPEAT(‘character-expression’,n) Repeats a character
particular condition that is true. expression n+1 times.
RIGHT(argument) Right-aligns a character expression
SAS Data Set Options ROUND(argument,round-off-unit) Rounds to the nearest
DROP=variable(s) Excludes variables from processing. round-off unit
FIRSTOBS=n Specifies the first observation to process SCAN(argument,n<,delimiters>) Returns a given word from
IN=variable Creates and names a variable that indicates a character expression
whether the data set contributed data to the current SUBSTR(argument,position<,n>) Extracts a substring from
observation. an argument.
KEEP=variable(s) Selects variables for processing. TIME() Returns the current time of day
LABEL='label' Specifies a label for a SAS data set TIMEPART(datetime) Extracts a time value from a SAS
OBS=n Specifies the first n observations to process datetime value
POINT=variable Direct observation number variable TODAY() Returns the current date as a SAS date value
RENAME=(oldname1=newname1 <...oldnamen=newnamen>) TRANSLATE(source,to,from) Replaces specific characters in
Changes the name of a variable. a character expression
WHERE=(expression1 <logical-operator expressionn>) SUM(argument,argument, ...) Returns the total value of the
Selects observations from a SAS data set that meet numeric arguments
certain conditions before SAS brings them into the DATA TRIM(argument) Takes the argument and removes any
or PROC step for processing. trailing blanks.
SAS Functions UPCASE(argument) Converts all letters in an argument to
BYTE(n) Returns one character in the ASCII or EBCDIC uppercase
collating sequence where n is an integer representing a WEEKDAY(date) Returns the day of the week from a SAS
specific ASCII or EBCDIC character date value
COMPBL(source) Removes multiple blanks from a character YEAR(date) Returns the year from a SAS date value
string SAS Formats
COMPRESS(source<,characters-to-remove>) w.d standard numeric
Removes specific characters from a character string COMMAw.d writes numeric values with commas and
DATE() Returns the current date as a SAS date value decimal points
DATEPART(datetime) Extracts the date from a SAS Zw.d print leading zeros
datetime value $w. writes standard character data
DATETIME() Returns the current date and time of day $CHARw. writes standard character data (including
DAY(date) Returns the day of the month from a SAS date leading blanks)
value $VARYINGw. Writes character data of varying length
HMS(hour,minute,second) Returns a SAS time value from
hour, minute, and second SAS Informats
INDEX(source,excerpt) Searches the source for the w.d Reads standard numeric data
character string specified by the excerpt datew. Reads date values (ddmmmyy)
LEFT(argument) Left-aligns a SAS character string $w. Reads standard character data
LENGTH(argument) Returns the length of an argument $VARYINGw. Reads character data of varying length
LOWCASE(argument) Converts all letters in an argument to Compliments of:
lowercase David Franklin
MAX(argument,argument, ...) Returns the largest value of New Hampshire, USA
the numeric arguments Tel/Fax +1(603) 262-9160 Cell +1(603) 275-6809
MDY(month,day,year) Returns a SAS date value from Email 100316.3451@compuserve.com
month, day, and year Google “Franklin SAS”
MIN(argument,argument, ...) Returns the smallest value of http://ourworld.compuserve.com/homepages/dfranklinuk
the numeric arguments Release 1.2b ©MMVII
 SAS® Cheat Sheet
SAS Procedures PROC FREQ <DATA=dset> PROC SORT <DATA=dset> <OUT=dset>
<ORDER=DATA|EXTERNAL|FREQ|INTERNAL>; <NODUPKEY|NODUPS>;
PROC COMPARE <BASE=dset> <COMPARE=dset>;
BY <DESCENDING> varn; BY <DESCENDING> variable-list;
BY variable(s);
TABLES requests </tables-options>; PROC TRANSPOSE <DATA=dset> <OUT=dset>;
ID variable(s);
where BY <DESCENDING> variable-list;
VAR variable(s);
requests one or more variable names joined by ID variable;
PROC DATASETS <LIBRARY=libref> <MEMTYPE=(m-list)> asterisks that specify the form of the VAR variable1 … variablen;
<DETAILS|NODETAILS> <KILL> generated tables, e.g. A*B
<NOLIST>; tables-options Can be one or more of the following: Macro Language
APPEND BASE=dset <DATA=dset> <FORCE>; LIST MISSING %DO macro-var=start_value %TO end_value <%BY step>;
CHANGE old-namen=new-namen </MEMTYPE=(m-list); NOPRINT OUT=SAS-data-set Executes a section of a macro repetitively based on the
CONTENTS <DATA=<libref.>member> <DIRECTORY> OUTPCT SPARSE value of an index variable
<MEMTYPE=(m-list)> <NODS> %DO %WHILE (expression);
PROC MEANS <DATA=dset> <DESCENDING>
<VARNUM> <NOPRINT> <OUT=dset>; Executes a section of a macro repetitively while a condition
<MISSING> <NOPRINT> <NWAY>
COPY OUT=libref <IN=libref> <MEMTYPE=(m-list)> is true
<ORDER=DATA|EXTERNAL|FREQ|
<MOVE>; %DO %UNTIL (expression);
INTERNAL>
EXCLUDE member-list </MEMTYPE=mtype>; Executes a section of a macro repetitively until a condition
<statistic-list>;
SELECT member-list </MEMTYPE=mtype>; is true
VAR variable-list;
DELETE member-list </ MEMTYPE=mtype>; %GLOBAL macro-variable(s);
CLASS variable-list;
MODIFY member-name <(<LABEL=’data-set-label’|’ ‘> Creates macro variables that are available during the
OUTPUT <OUT=dset> <out-statistic>;
<SORTEDBY=sort-information>)>; execution of an entire SAS session
where
FORMAT variable-format-name.; %IF expression %THEN action; <%ELSE action;>
statistic-list Can be one or more of the following:
INDEX CREATE variable </<UNIQUE> <NOMISS>>; Conditionally process a portion of a macro
N, NMISS, MIN, MAX, RANGE,
INDEX CREATE index=(variable-list) </<UNIQUE> %LENGTH (character string | text expression)
MEDIAN, SUM, MEAN, VAR, STD,
<NOMISS>>; Returns the length of a string
Q1, Q3, T
INDEX DELETE index-list; %LET macro-variable =<value>;
out-statistic Specifies the statistics in the output and
LABEL variable=’label-text’; Creates a macro variable and assigns it a value
also names the variable(s) that contain
RENAME variablen=new-variablen; %MACRO m-name (<pp1><…,ppn><kp1=value<..<kpn=v>);
the results.
QUIT; Begins a macro definition
where PROC REPORT <DATA=dset> <HEADLINE> <HEADSKIP>
%MEND <macro-name>;
m-list one or more of the member types that <NOWINDOWS> <SPACING=number>;
Ends a macro definition
processing should be restricted to. COLUMNS <report-item1, <., report-itemn>>
%SCAN(argumentn<,delimiters>)
member-list list of members in the directory to (`header1 ' < . `headern '> report-item(s) );
Search for a word that is specified by its position in a string
process. DEFINE report-item / <usage> <define-options>;
%SUBSTR(argument,position<,length>)
mtype restricts processing to one member type. COMPUTE <BEFORE|AFTER> report-item;
Produce a substring of a character string
LINE <item item-format | 'text' | pointer-control>;
PROC EXPORT DATA=<libref.>dset %UPCASE(character string | text expression)
ENDCOMP;
OUTFILE="filename" <REPLACE>; Convert values to uppercase
BREAK BEFORE|AFTER break-variable </b-option(s)>;
PROC IMPORT DATAFILE="filename" Macro Quoting
QUIT;
OUT=<libref.>dset <REPLACE>; %QUOTE | %NRQUOTE and %BQUOTE | %NRBQUOTE
where
The following filetypes are the most commonly used and Mask special characters and mnemonic operators in a
report-item name or alias (established in the
supported within filename by SAS: resolved value at macro execution
COLUMN statement) of the data set or
filename.XLS (Microsoft Excel) %STR | %NRSTR
computed variable, or statistic to define
filename.TXT (tab delimited) Mask special characters and mnemonic operators in
usage Either ACROSS, ANALYSIS,
filename.CSV (comma separated value) constant text at macro compilation
COMPUTED, DISPLAY, GROUP,
PROC FORMAT <CNTLIN=dset> %SUPERQ
ORDER
<CNTLOUT=dset> Masks special characters/mnemonic operators at macro
define-options The following options are available:
<LIBRARY=libref<.catalog>>; execution but prevents further resolution of the value.
FORMAT=format ORDER=
INVALUE <$>name <value-range-set(s)>; SPACING= WIDTH= Compliments of:
PICTURE name <...value-range-set <(picture-option(s))>>; DESCENDING FLOW David Franklin
VALUE <$>name <value-range-set>; NOPRINT CENTER New Hampshire, USA
where LEFT RIGHT Tel/Fax +1(603) 262-9160 Cell +1(603) 275-6809
picture-options The following options are useful: COLOR= ‘column-header’ Email 100316.3451@compuserve.com
ROUND NOEDIT b-options These include: Google “Franklin SAS”
PREFIX= FILL= SKIP PAGE http://ourworld.compuserve.com/homepages/dfranklinuk
Release 1.2b ©MMVII