Missing data

As R was designed to analyze datasets, it includes the concept of missing data (which is uncommon in other programming languages). Missing data are represented in vectors as NA.

When doing operations on numbers, most functions will return NA if the data you are working with include missing values. This feature makes it harder to overlook the cases where you are dealing with missing data. You can add the argument na.rm=TRUE to calculate the result while ignoring the missing values.

rooms <-c(2, 1, 1, NA, 4) 
mean(rooms) 
[1] NA 
max(rooms) 
[1] NA 
mean(rooms, na.rm = TRUE) 
[1] 2 
max(rooms, na.rm = TRUE) 
[1] 4 

If your data include missing values, you may want to become familiar with the functions is.na(), na.omit(), and complete.cases(). See below for examples.

## Extract those elements which are not missing values. 
rooms[!is.na(rooms)] 
[1] 2 1 1 4 
## Count the number of missing values. 
sum(is.na(rooms)) 
[1] 1 
## Returns the object with incomplete cases removed. The returned object is an atomic vector of type `"numeric"` (or `"double"`). 
na.omit(rooms) 
[1] 2 1 1 4 
attr(,"na.action") 
[1] 4 
attr(,"class") 
[1] "omit" 
## Extract those elements which are complete cases. The returned object is an atomic vector of type `"numeric"` (or `"double"`). 
rooms[complete.cases(rooms)] 
[1] 2 1 1 4 

Exercise

Using this vector of rooms, create a new vector with the NAs removed.
```
rooms <-c(1, 2, 1, 1, NA, 3, 1, 3, 2, 1, 1, 8, 3, 1, NA, 1) 
```
Use the function median() to calculate the median of the rooms vector.
Use R to figure out how many households in the set use more than 2 rooms for sleeping.