/* Three ways to read excel files into SAS */
/* the default format is assumed to be:
/*   row 1 = variable names,
/*   subsequent rows = observations, 
/*   columns = variables */

/* 1) from a .xls file */
proc import datafile='c:/philip/stat 402/camp.xls' out=camp
  replace;
run;
/* replace tells SAS to reread the file if the data set already */
/*   exists.  This is necessary if you've corrected the data file */

/* additional options (place before the ;) are:
   sheet = 'name'  to read a specific sheet from a workbook
     not needed if the sheet is named sheet1 (the excel default)
   getnames = no   when the first row is data, not variable names */

/* the file must NOT be open in excel, otherwise SAS 
/*  refuses to read the file */

proc print data=camp(obs=10);
title 'Importing a .xls file';
run;

/*2)  from a .csv copy of a specific sheet */
proc import datafile='c:/philip/stat 402/camp.csv' out=camp
  replace;
/* the file must NOT be open in excel, otherwise SAS 
/*  refuses to read the file */

proc print data=camp(obs=10);
title 'Importing a .csv file ';
run;

/* if you want to do any transformations, computations,
/*   or recoding, then you need another data step */

data camp2;
  set camp;
  loghr0 = log(hr0);
  loghr4 = log(hr4);


/*3) using a data step and input statement to read a .csv file */ 
data camp;
  infile 'c:/philip/stat 402/camp.csv' dsd firstobs=2;
  input .....;

  /* put additional data step commands here */

/*  you need to provide the variable names in the input command */

/* the dsd option to the infile is crucial */
/*   it tells SAS to deal with commas and quotes in the appropriate
/*   way  to correctly read a .csv file */
/* the firstobs=2 tells SAS to skip the first line */
/*   appropriate only if the first line is column names */

proc print data=camp(obs=10);
title 'Using a data step';
run;

/*4)  reading a tab-delimited file 
/*  (with tab characters instead of spaces between values )  */

data camp;
  infile 'c:/philip/stat 402/camp.prn' dlm='09'x;
  input .....;

/* the dlm= changes the delimiter, the separator between */
/*  fields.  The '09'x is the tab character in hexadecimal code */
 
/* And finally, how to save a SAS data set as an excel file */
/* imagine that SAS has created a data set called means */
proc export data=means outfile='c:\philip\stat 402\means.xls';
/* to create an excel workbook */
run;

proc export data=means outfile='c:\philip\stat 402\means.csv';
/* to create a .csv file */
run;