The `aggregate()`

function subsets dataframes, and time series data, then computes summary statistics. The structure of the `aggregate()`

function is `aggregate(x, by, FUN)`

.

Answers to the exercises are available here.

**Exercise 1**

Aggregate the “`airquality`

” data by “`airquality$Month`

“, returning means on each of the numeric variables. Also, remove “`NA`

” values.

**Exercise 2**

Aggregate the “`airquality`

” data by the variable “`Day`

“, remove “`NA`

” values, and return means on each of the numeric variables.

**Exercise 3**

Aggregate “`airquality$Solar.R`

” by “`Month`

“, returning means of “`Solar.R`

“. The header of column 1 should be “`Month`

“. Remove “`not available`

” values.

**Exercise 4**

Apply the standard deviation function to the data aggregation from Exercise 3.

**Exercise 5**

The structure of the `aggregate()`

formula interface is `aggregate(formula, data, FUN)`

.

The structure of the formula is `y ~ x`

. The “`y`

” variables are numeric data. The “`x`

” variables, usually factors, are grouping variables, that subset the “`y`

” variables.

`aggregate.formula`

allows for one-to-one, one-to-many, many-to-one, and many-to-many aggregation.

Therefore, use `aggregate.formula`

for a one-to-one aggregation of “`airquality`

” by the mean of “`Ozone`

” to the grouping variable “`Day`

“.

**Exercise 6**

Use `aggregate.formula`

for a many-to-one aggregation of “`airquality`

” by the mean of “`Solar.R`

” and “`Ozone`

” by grouping variable, “`Month`

“.

**Exercise 7**

Dot notation can replace the “`y`

” or “`x`

” variables in `aggregate.formula`

. Therefore, use “`.`

” dot notation to find the means of the numeric variables in `airquality`

“, with the grouping variable of “`Month`

“.

**Exercise 8**

Use dot notation to find the means of the “`airquality`

” variables, with the grouping variables of “`Day`

” and “`Month`

“. Display only the first 6 resulting observations.

**Exercise 9**

Use dot notation to find the means of “`Temp`

“, with the remaining “`airquality`

” variables as grouping variables.

**Exercise 10**

`aggregate.ts`

is the time series method for `aggregate()`

.

Using `R`

‘s built-in time series dataset, “`AirPassengers`

“, compute the average annual standard deviation.

Image by Averater (Own work) [CC BY-SA 3.0], via Wikimedia Commons.

LEAA says

Very helpful! Thanks a lot.

John Akwei says

Thank you for your compliment!

Huw says

In excercise 10, why is aggregate.ts given as a hint if the explanation is to not use it?

Onno Dijt says

Hi Huw,

To answer Exercise 10, you can use both aggregate and aggregate.ts, it is up to you to decide!

Have fun!

Huw says

Thank you!

Huw says

Looking at it, is there a functional difference between:

aggregate(airquality$Ozone, list(airquality$Day), mean, na.rm=TRUE)

and

aggregate(Ozone ~ Day, airquality, mean)

It seems to be that rather than including the ‘list(airquality)’ section, it would be easier to do everything as X~Y and use dot notation but I’m not sure if I’m missing something in the background.

John Akwei says

Yes. “aggregate(airquality$Ozone, list(airquality$Day), mean, na.rm=TRUE) ” was not a suggested answer, because the Exercise referred to the formula interface of aggregate().

Jan Trommelmans says

Exercise 10 asks for the average annual standard deviation. Should the solution not be: mean(aggregate(AirPassengers,nfrequency=1,sd)) ?

John Akwei says

Thanks for inquiring, Jan!

I believe I was referring to aggregate() finding the means of a set of observations.

However, I will possible modify the question.

Best Regards,

John Akwei, Data Scientist