White Noise Time Series with Python

White noise is an important concept in time series forecasting.

If a time series is white noise, it is a sequence of random numbers and cannot be predicted. If the series of forecast errors are not white noise, it suggests improvements could be made to the predictive model.

In this tutorial, you will discover white noise time series with python.

After completing this tutorial, you will know:

The definition of a white noise time series and why it matters. How to check if your time series is white noise. Statistics and diagnostic plots to identify white noise in Python.

Let’s get started.

White Noise Time Series with Python

Photo by Dan Eckert , some rights reserved.

What is a White Noise Time Series?

A time series may be white noise.

A time series is white noise if the variables are independent and identically distributed with a mean of zero.

This means that all variables have the same variance ( sigma^2 ) and each value has a zero correlation with all other values in the series.

If the variables in the series are drawn from a Gaussian distribution, the series is called Gaussian white noise.

Why Does it Matter?

White noise is an important concept in time series analysis and forecasting.

It is important for two main reasons:

Predictability : If your time series is white noise, then, by definition, it is random. You cannot reasonably model it and make predictions. Model Diagnostics : The series of errors from a time series forecast model should ideally be white noise.

Model Diagnostics is an important area of time series forecasting.

Time series data are expected to contain some white noise component on top of the signal generated by the underlying process.

For example:

y(t) = signal(t) + noise(t)

Once predictions have been made by a time series forecast model, they can be collected and analyzed. The series of forecast errors should ideally be white noise.

When forecast errors are white noise, it means that all of the signal information in the time series has been harnessed by the model in order to make predictions. All that is left is the random fluctuations that cannot be modeled.

A sign that model predictions are not white noise is an indication that further improvements to the forecast model may be possible.

Stop learning Time Series Forecasting the slow way Sign-up and get a FREE 7-day Time Series Forecasting Mini-Course

You will get:

... onelesson each day delivered to your inbox

... exclusive PDF ebook containing all lessons

...

confidence and skills

to work through your own projects

Download Your FREE Mini-Course

Is your Time Series White Noise?

Your time series is not white noise if any of the following conditions are true:

Does your series have a zero mean? Does the variance change over time? Do values correlate with lag values?

Some tools that you can use to check if your time series is white noise are:

Create a line plot . Check for gross features like a changing mean, variance, or obvious relationship between lagged variables. Calculate summary statistics . Check the mean and variance of the whole series against the mean and variance of meaningful contiguous blocks of values in the series (e.g. days, months, or years). Create an autocorrelation plot . Check for gross correlation between lagged variables. Example of White Noise Time Series

In this section, we will create a Gaussian white noise series in Python and perform some checks.

It is helpful to create and review a white noise time series in practice. It will provide the frame of reference and example plots and statistical tests to use and compare on your own time series projects to check if they are white noise.

Firstly, we can create a list of 1,000 random Gaussian variables using the gauss() function from the random module .

We will draw variables from a Gaussian distribution with a mean ( mu ) of 0.0 and a standard deviation ( sigma ) of 1.0.

Once created, we can wrap the list in a Pandas Series for convenience.

fromrandomimportgauss fromrandomimportseed frompandasimportSeries frompandas.tools.plottingimportautocorrelation_plot # seed random number generator seed(1) # create white noise series series = [gauss(0.0, 1.0) for i in range(1000)] series = Series(series)

Next, we can calculate and print some summary statistics, including the mean and standard deviation of the series.

# summary stats print(series.describe())

Given that we defined the mean and standard deviation when drawing the random numbers, there should be no surprises.

count1000.000000 mean -0.013222 std 1.003685 min-2.961214 25%-0.684192 50%-0.010934 75% 0.703915 max 2.737260

We can see that the mean is nearly 0.0 and the standard deviation is nearly 1.0. Some variance is expected given the small size of the sample.

If we had more data, it might be more interesting to split the series in half and calculate and compare the summary statistics for each half. We would expect to see a similar mean and standard deviation for each sub-series.

Now we can create some plots, starting with a line plot of the series.

# line plot series.plot() pyplot.show()

We can see that it does appear that the series is random.

White Noise Series Line Plot

We can also create a histogram and confirm the distribution is Gaussian.

# histogram plot series.hist() pyplot.show()

Indeed, the histogram shows the tell-tale bell-curve shape.

White Noise Series Histogram Plot

Finally, we can create a correlogram and check for any autocorrelation with lag variables.

# autocorrelation autocorrelation_plot(series) pyplot.show()

The correlogram does not show any obvious autocorrelation pattern.

There are some spikes above the 95% and 99% confidence level, but these are a statistical fluke.

White Noise Series Correlogram Plot

For completeness, the complete code listing is provided below.

fromrandomimportgauss fromrandomimportseed frompandasimportSeries frompandas.tools.plottingimportautocorrelation_plot frommatplotlibimportpyplot # seed random number generator seed(1) # create white noise series series = [gauss(0.0, 1.0) for i in range(1000)] series = Series(series) # summary stats print(series.describe()) # line plot series.plot() pyplot.show() # histogram plot series.hist() pyplot.show() # autocorrelation autocorrelation_plot(series) pyplot.show() Further Reading This section lists some resources for furthe