Lagging and differencing

1. Lagging and differencing

Another important component of time series data manipulation is generating moving indicators from your data. Two common types of moving indicators are lags and differences.

2. Lagging

As you may recall from the previous course, a lag offsets the value of your time series data by a certain number of periods. This is an extremely useful metric for time series analysis, especially when you think the effect of one variable takes a while to set in. For example, if you're interested in how economic trends influence the tourism industry, you should keep in mind that tourists often plan their trips far in advance. In this context, it makes sense to lag our data to allow economic changes to have an effect.

3. Differencing

A time series difference takes this idea one step further. Rather than simply taking the value of your time series at each period, a difference tells you how much your measure has changed since some prior period. If you're interested in economic growth, it may make sense to focus on the difference in unemployment from one period to another, rather than values at any give time.

4. Let's practice!

In the next few exercises, you'll generate some useful lags and difference values in your economic data. OK, let's get to work!

Create Your Free Account

or

By continuing, you accept our Terms of Use, our Privacy Policy and that your data is stored in the USA.

This exercise is part of the course

Case Study: Analyzing City Time Series Data in R

IntermediateSkill Level

4.9+

Start Course for Free

You've been hired to understand the travel needs of tourists visiting the Boston area. As your first assignment on the job, you'll practice the skills you've learned for time series data manipulation in R by exploring data on flights arriving at Boston's Logan International Airport (BOS) using xts & zoo.

Exercise 1: Review xts fundamentals Exercise 2: Identify the time series Exercise 3: Flight data Exercise 4: Pick out the xts object Exercise 5: Encoding your flight data Exercise 6: Manipulating and visualizing your data Exercise 7: Exploring your flight data Exercise 8: Visualize flight data Exercise 9: Calculate time series trends Exercise 10: Saving and exporting xts objects Exercise 11: Assessing flight trends Exercise 12: Saving time - I Exercise 13: Saving time - II

In this chapter, you'll expand your time series data library to include weather data in the Boston area. Before you can conduct any analysis, you'll need to do some data manipulation, including merging multiple xts objects and isolating certain periods of the data. It's a great opportunity for more practice!

Exercise 1: Merging time series data by row Exercise 2: Exploring temperature data Exercise 3: Next steps - I Exercise 4: Merging using rbind()Exercise 5: Visualizing Boston winters Exercise 6: Merging time series data by column Exercise 7: Subsetting and adjusting periodicity Exercise 8: Generating a monthly average Exercise 9: Using merge() and plotting over time Exercise 10: Are flight delays related to temperature?Exercise 11: Time series data workflow Exercise 12: Next steps - II Exercise 13: Expanding your data Exercise 14: Are flight delays related to visibility or wind?

Now it's time to go further afield. In addition to flight delays, your client is interested in how Boston's tourism industry is affected by economic trends. You'll need to manipulate some time series data on economic indicators, including GDP per capita and unemployment in the United States in general and Massachusetts (MA) in particular.

Exercise 1: Handling missingness Exercise 2: Exploring economic data Exercise 3: Replace missing data - I Exercise 4: Replace missing data - II Exercise 5: Estimating missing GDP Exercise 6: Lagging and differencing

Current Exercise

Exercise 7: Exploring unemployment data Exercise 8: Lagging unemployment Exercise 9: Differencing unemployment Exercise 10: Rolling functions Exercise 11: Add a discrete rolling sum to GDP data Exercise 12: Add a continuous rolling average to unemployment data Exercise 13: Manipulating MA unemployment data

Having exhausted other options, your client now believes Boston's tourism industry must be related to the success of local sports teams. In your final task on this project, your supervisor has asked you to assemble some time series data on Boston's sports teams over the past few years.

Exercise 1: Advanced features of xts Exercise 2: Encoding and plotting Red Sox data Exercise 3: Calculate a closing average Exercise 4: Calculate and plot a seasonal average Exercise 5: Calculate and plot a rolling average Exercise 6: Indexing commands in xts Exercise 7: Extract weekend games Exercise 8: Calculate a rolling average across all sports Exercise 9: Viewing sports trends Exercise 10: Congratulations