How to create time series in r

Intro

Time series is one of the most common analysis and modeling in Data Science. In this article, we will learn how to create time series in R.

Creating a Basic Time Series

Let’s say we had a vector of sales data. In this example, we have five elements with dollars values, i.e. $300,$400, etc.

sales <- c(300, 400, 100, 400, 800)

Now, let’s say we want to convert this to a time series object which will index our data by date. We can use the built in ts method. We will use three parameters: data, which will take our sales data above, start, which will set to start In January of 2018, and frequency, which will set to 1 to represent once a year.

sales.data <- ts(data = sales, start = c(2018, 1), frequency = 1)
sales.data

## Time Series:
## Start = 2018 
## End = 2022 
## Frequency = 1 
## [1] 300 400 100 400 800

Here you can see that we now have a Time Series object that spans from 2018 to 2022.

If our data was not yearly, we can change the frequency. Let’s assume the sales were quartly. We can change the frequency to 4.

sales.data <- ts(data = sales, start = c(2018, 1), frequency = 5)
sales.data

## Time Series:
## Start = c(2018, 1) 
## End = c(2018, 5) 
## Frequency = 5 
## [1] 300 400 100 400 800

Now our data represents sales per a quarter.

The Zoo Library

The built in ts function is great, but there are alternatives that you may see while using time series data.

Let’s start with the zoo library, which adds a lot of functionality to the build in time series. We can create an object using the zoo function.

library(zoo)

## Warning: package 'zoo' was built under R version 4.0.5

## 
## Attaching package: 'zoo'

## The following objects are masked from 'package:base':
## 
##     as.Date, as.Date.numeric

zoo(sales.data)

##   2018 2018.2 2018.4 2018.6 2018.8 
##    300    400    100    400    800

If we didn’t have a time series object already, we can pass our sales data and a sequence of dates to zoo.

days <- seq(as.Date("2018-01-01"), as.Date("2018-01-05"), by = "days")

zoo(sales, days)

## 2018-01-01 2018-01-02 2018-01-03 2018-01-04 2018-01-05 
##        300        400        100        400        800

The xts library

Another library you may see is the xts library. It extends zoo, so anything that works with zoo also works with xts. We can create an xts time series the same way as zoo.

library(xts)

## Warning: package 'xts' was built under R version 4.0.5

days <- seq(as.Date("2018-01-01"), as.Date("2018-01-05"), by = "days")

xts(sales, days)

##            [,1]
## 2018-01-01  300
## 2018-01-02  400
## 2018-01-03  100
## 2018-01-04  400
## 2018-01-05  800