statgeek

/*------------------------------------------------------------------*
| MACRO NAME  : tablen
| SHORT DESC  : Creates a descriptives table of multiple variables
*------------------------------------------------------------------*
| CREATED BY  : Meyers, Jeffrey                 (07/28/2016 3:00)
*------------------------------------------------------------------*
| VERSION UPDATES:
| 2.46 11/03/2021
|   Changed code to avoid truncating character strings at 200
| 2.45 08/25/2021

statgeek

*path to the zip file;
filename src zip "/home/fkhurshed/Demo2/P_DR2IFF.zip";

*path to where to save the xpt file;
filename xl "/home/fkhurshed/Demo2/P_DR2IFF.xpt" ;
 
*extract file from zip - P_DR2IFF.XPT in the code below is the name of the file in the zipped file that is to be extracted;
data _null_;
   /* using member syntax here */
   infile src(P_DR2IFF.XPT) 

statgeek

/*Author: F.Khurshed
Date: 2022-10-07*/


*create library to store format catalog;
libname demo '/home/fkhurshed/Demo1';

proc format;
value age_fmt 
  low - 12 = 'Child'

statgeek

/*This program illustrates how to loop through by calendar months and calculate a moving statistic*/

data have;
set sashelp.stocks;
run;

*sort for faster processing and add index;
proc sort data=have out=have (index=(date));
by date;
run;

statgeek

*This macro will export a file to a data set and split it based on the number of records per sheet;
%macro export_split (dsn=, size=);
    %*Get number of records and calculate the number of files needed;
    data _null_;
        set &dsn. nobs=_nobs;
        call symputx('nrecs', _nobs);
        n_files=ceil(_nobs/ &size.);
        call symputx('nfiles', n_files);
        stop;
    run;

statgeek

/*This is an example of how to create an Other category for everything except the top 3*/
*not ideal, other becomes first format - will look into how to make it last value;

*get counts;
proc freq data=sashelp.class order = freq;
table age / out=counts;
run;

*create format;
data counts_fmt;

statgeek

*This example demonstrates how you can merge data with a formatted variable and not have to create a new variable;

data stocks_A;
set sashelp.stocks;
where stock='IBM';
format date yymmn6.;
*keep only relevant variables for testing;
keep date open;

*rename to identify source;

statgeek

*How to selectively filter your list of variables in a SAS data set with a pattern but not one that uses SAS variable shortcuts;

 options mprint symbolgen;

%macro select(lib =, ds_in=, pattern=, ds_out=);

proc sql noprint;
select nliteral(name) into :var_list separated by ' '
from dictionary.columns
where libname = upcase("&lib")

statgeek

/*This is an example of how to search through a list of terms and see if a field contains any of the values*/

*Make fake data to show example;
*terms to search for;
data terms; 
set sashelp.baseball (obs=5);
search_term = substr(team,1,3);
keep search_term;;
run;
    

statgeek

#This program centers all variables
#This is also called standardization - each variable minus the mean. 


library(tidyverse)


#generate fake data
df  <-  structure(list(day = c(2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 
2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,