← SASnRSubset observations and variables

Introduction

Real analysis rarely uses a raw dataset as-is — we usually filter to the rows we care about AND pick just the columns we need in the same step.
SAS does both in one DATA step: a WHERE clause for rows, plus keep or drop for columns.
In tidyverse we chain filter() and select() with the pipe. In base R we combine a logical row index and a column index in a single [ call.
This lesson keeps only the male subjects and drops the sex column in one step — a typical pattern when we subset to one treatment group and no longer need the grouping variable.
After running, we have a males dataset with only the male rows and no Sex column.

Subset observations and variables

SAS (Base SAS)+

data CLASS;
 infile datalines dlm='|' dsd missover;
input Name : $8. Sex : $1. Age : best32. Height : best32. Weight : best32.;
label ;
 format ;
 datalines4;
Alfred|M|14|69|112.5
Alice|F|13|56.5|84
Barbara|F|13|65.3|98
Carol|F|14|62.8|102.5
Henry|M|14|63.5|102.5
James|M|12|57.3|83
;;;;
 run;

data males;
    set class;
    where sex="M";
   drop sex;
 run;

To filter rows and drop columns in a single pass, we combine WHERE and drop inside one DATA step.
set class; reads the input. where sex="M"; restricts the incoming rows. drop sex; removes the sex column from the output.
Order matters conceptually: the WHERE filter runs as rows are read, and the DROP removes variables from the output buffer. So the sex value is available for the WHERE test even though it does not appear in the final dataset.
We could equally write keep name age height weight; — the result is identical. drop is handier when there are only a few columns to remove; keep is clearer when there are only a few to keep.
Output males has the male rows from CLASS and every column except sex.

R (tidyverse)+

library(tidyverse) 

class<-tribble( 
~Name,~Sex,~Age,~Height,~Weight, 
"Alfred","M",14,69,112.5, 
"Alice","F",13,56.5,84, 
"Barbara","F",13,65.3,98, 
"Carol","F",14,62.8,102.5, 
"Henry","M",14,63.5,102.5, 
"James","M",12,57.3,83, ) 

males<-class %>% filter(Sex=="M") %>% select(-Sex)

In tidyverse we pipe the tibble through filter() and then select().
class %>% filter(Sex == "M") %>% select(-Sex) reads as three steps: start with class, keep rows where Sex is M, then drop the Sex column.
The pipe operator %>% passes the output of each step as the first argument to the next, so we do not have to name intermediate tibbles.
The order of filter() and select() does not affect the result in this simple case, but filter-then-select is usually clearer because we focus on rows first and worry about columns afterwards.
The result is a tibble with only the male rows and no Sex column.

Happy learning!

Search lessons

Introduction

Subset observations and variables