Practical Data Management and Statistical Computing (BioEp691F) Contacts Outline Assignments Resources Outline Outline: Lec1 Lec2 Lec3 Lec4 Lec5 Lec6 Lec7 Lec8 Lec9 Lec10 Lectures: Lec1 Lec2 Lec3 Lec4 Lec5 Lec6 Lec7 Lec8 Lec9 Lec10 Outline: Lec11 Lec12 Lec13 lec14 Lec15 Lec16 Lec17 Lec18 Lec19 Lec20 Lectures: Lec11 Lec12 Lec13 Lec14 Lec15 Lec16 Lec17 Lec18 Lec19 Lec20 Outline: Lec21 Lec22 Lec23 lec24 Lec25 Lectures: Lec21 Lec22 Lec23 Lec24 Lec25 Lecture 1 Scope of Course Course Expectations Computers Background Survey Example 1. Weight Change Program Reading data with INPUT statement with mixed format Printing SAS data set Using PROC TABULATE to get means by group Lecture 2 SAS STATEMENT keywords DATA INPUT CARDS or DATALINES RUN Variable Names in SAS Adding Comments: * Adding Buttons to control processing. Adding Titles TITLE1 "Source: es99p1.sas 9/14/99"; TITLE2 "Mean weight at various times by group"; Saving programs and data Using a previously created SAS data set. SET Creating a permanent SAS data set. LIBNAME Example: LIBNAME new 'c:\temp\'; DATA new.lec1; Documenting and Naming SAS programs with a Common Header the OPTIONS statement the Project Name the TITLE1 statement (with program name, programmer, date) the Description data files read data files created the LIBNAME statements Lecture 3 Review Using a SAS data set again in the same SAS Session SET Creating a Permanent SAS Data set with a LABEL. LIBNAME new 'c:\temp\'; DATA new.lec1 (LABEL="desc"); Getting the Contents of the data PROC CONTENTS Reading Data from a Previously Saved ASCII Data File INFILE Statement Downloading ASCII data from the WEB and stripping special characters Identifying special ASCII characters in an ASCII data file. Reading Data in Column Input. Determining columns for variables in a data set INFILE Statement with options: CARDS OBS= FIRSTOBS= LIST statement Lecture 4 Review Downloading ASCII data from the WEB Reading data from the ICU study. Identifying ASCII codes in a data set INFILE with the option MISSOVER (to go to a new record) INPUT (v1-v5) ($1.) ; as a shorthand to input a set of variables ARRAY v{5} ; to define an array of variables DO i=1 TO 5; ...... END; to perform an operation on variables FILE PRINT; to route output from PUT statements to the OUTPUT window PUT ..... to write out results directly as data lines are processed. SAS Functions BYTE(v1) to convert an ASCII code to the ASCII character. RANK(v1) to determine the ASCII code equivalent of an ASCII character variable Using ANALYZE to create a description of data Lecture 5 Using PROC PRINT Basic Printing Tasks Options to PROC PRINT DATA= (OBS= ) NOOBS Optional Statements: VAR WHERE ID Special Printing Tasks Options to PROC PRINT SPLIT=" /" ROWS=PAGE Optional Statements BY (with PROC SORT) SUM SUMBY Making Lists easy to read Options to PROC PRINT HEADING=V WIDTH=U Naming Values for Categorical Variables (PROC FORMAT) Creating format names with PROC FORMAT Optional Statements VALUE Linking format names to variables (FORMAT) With PROC PRINT With PROC FREQ Using PROC FREQ Adding Format names Required TABLES Statement Optional Statements NOROW NOCOL NOCUM NOPERCENT MISSING MISSPRINT LIST   Lecture 6 Using PROC CHART Using PROC MEANS Using PROC PLOT Using PROC UNIVARIATE Using PROC GCHART Lecture 7 Using Endnote 3.0 for Bibliography Management Construction of a reference list with Endnote Working with references Creating a bibliography for a research paper (using MS Word) Searching remote citation data bases Lecture 8 Making a Web Site Using Claris HOMEPAGE Overview Basics Initial Upload at UMass Password Protection Lecture 9 Reading ASCII data with more than one line per record use of / or #1 check for matching duplicate variables check for properly ordered lines using a subset of variables with a KEEP option in a DATA statement. More on PROC FORMAT Saving permanent Formats Using CNTLOUT and CNTLIN options. Defining formats for character variables ($fmtname) Lecture 10 Alternative Ways to store PROC FORMATS use of CNTLOUT and CNTLIN use of automatic LIBRARY="libref" option automatically generates a Format Catalogue called FORMATS.SC2 uses the same name for the catalogue for all projects use FMTSEARCH=(libref) in OPTIONS statement to use the formats use NOMFTERR in OPTIONS statement to override unknown formats saved in a SAS data set. Reading data with unequal # lines per subject- introduction Lecture 11 Creating Multiple Data sets in a DATA step OUTPUT statement Automatic FIRST.xxx and LAST.xxx variables RETAIN to keep values to use in another record Working with Data in a DATA step concatenating two SAS data sets with a SET statement using a match MERGE statement with a BY variables High Resolution Graphics SYMBOL# statement I=JOIN to connect points C=BLACK to select color R=100 to repeat for more SYMBOL# statements PROC GPLOT PLOT x*y=z to overlay plots by z option NOLEGEND to supress the legend for subjects Lecture 12 Automating DOS and Windows Commands in SAS Use X , return with EXIT Reading older SAS data sets with Engines LIBNAME V604 old4 'c:\data'; Locating ASCII data FILENAME statement INFILE statement More on LIST input deliminators reading variables with an embedded blank Lecture 13 List Input Problems Missing Data: INFILE options MISSOVER, LRECL Embedded blanks: & Long character Variables: $12. Column Input Options Pointer Control: +3 Multiple lines per record: #1, #2, or / Control of Columns: @16 INFILE Options DELIMITER=","; INFORMATS Dates Lecture 14 More on DATE INFORMATS DATE8. formats (with month names) MMDDYY8. formats (all numbers) Year 2000 Issues Setting the OPTION YEARCUTOFF=1920; Creating age from birthdates Other INFORMATs Reading dollar amounts: COMMAxx.x Reading addresses and phone numbers Using LENGTH statements Using Character functions LENGTH function (counts non-blank characters) SCAN function (identify words) TRIM function (deletes trailing blanks) LEFT function (left justifies a string) Writing an Address List the FILE statement Special SAS KEYWORDS: _NULL_ a data set that contains nothing _N_ a counter for each consecutive records in a data step Using PUT statements Lecture 15 Using LENGTH statements for character variables use to prevent the first occurance of a new character variable setting to small a character length Using LENGTH statements for numeric variables (and dangers) generally do not set length for numeric variables- allow them to have length 8. Lecture 16 Reading DATA saved by other Data Base Systems Access files *.sa2 View Files *.sv2 Reading DBASE files with the IMPORT wizzard Reading EXCEL files with the IMPORT wizzard Using DBMS COPY to translate files Reading Pipe Deliminated files through MS ACCESS to EXCEL to SAS Lecture 17  Macro Variables Defining Numeric: % LET a=25; Using: TITLE2 "Study with &a subjects"; INPUT statement with Trailing @ to hold the line for next INPUT SELECT ; END; statements INPUT statement with Trailing @@ to hold line for repeated input Options that make Debugging SAS programs Easier Using the LIBRARY ICON Using KEEP and DROP in DATA/SET statemtents USING SPLIT="*" with PROC PRINT, and FORMATS More on SAS VIEWS Lecture 18 Handling MISSING VALUES in SAS Recogonizing Missing and Invalid Numeric Values Using OPTION INVALIDDATE="Z" to assign all characters to the special missing value .Z Recogonizing Missing and Invalid Character Variables Converting Numberic and Character Variables Converting character values to numeric values Details on Storage of numeric variables on PCs and in SAS Recoding missing value codes to SAS missing values and other Special Missing Values Lecture 19  Using EPI INFO to enter and verify data Lecture 20 Combining and Manipulating Data Overview DATA Step Options: SET, MERGE, UPDATE PROCedure Options: APPEND, COPY, DATASETS, TRANSPOSE Example: 5 women measured at three times resulting in 3 data sets Concatenating SAS data sets Problems with different variable names for different times. Pre-sorting to InterLeave SAS data sets: BY ID; Using RETAIN to keep a variable for subsequent records Using FIRST. and LAST. variables with BY statement. Using Automatic IN variables to control processing Merging SAS Data Sets Sort and MATCH merge BY statement Avoid overlapping variables: last data set takes priority (order matters) Duplicate records Merging by more than one variable Using IN variables in a merge Using FIRST. and LAST. variables in a MERGE Lecture 21 Creating a Codebook Creating a WEB site to Document Data Lecture 22 Selecting a simple random sample with replacement  Using a Uniform Random Number Generator Selecting a simple random sample without replacement Using a RETAIN statement Using OUTPUT statements Using PUT statements to see how it works. Selecting a stratified simple random sample Lecture 23 Simulations (Illustrating the central limit theorem) SAS random number generators Controlling the random numbers with (SEED) Using a CALL statement to control several strings of random numbers Using RANBIN to generate binomial random numbers Using RANNOR to generate normal (0,1) random variables Evaluating the Central Limit Theorem with n=7 Evaluating the Central Limit Theorem with n=30 Using MACRO variables the Simulation. Lecture 24  Simulations (Evaluating how good the Relative Odds ratio is to the Risk Ratio in studies) The Problem The Cohort Study The Case Control Study Writing a MACRO program %MACRO name (a,b); %MEND name; %name(2,5) A simple example Designing the simulation Using Arrays Conducting the Simulation Using MACROs to help Lecture 25

Produced and maintained by the Dept of BioEpi at UMASS Send comments or questions about this web site to Ed Stanek Email: stanek@schoolph.umass.edu \be691f\web\webready\outline.html Lst Update: 10/27/99