1. Learn
    2. /
  2. Courses
    3. /
  3. Visualizing Time Series Data in Python

Exercise

Display rolling averages

It is also possible to visualize rolling averages of the values in your time series. This is equivalent to "smoothing" your data, and can be particularly useful when your time series contains a lot of noise or outliers. For a given DataFrame df, you can obtain the rolling average of the time series by using the command:

df_mean = df.rolling(window=12).mean()

The window parameter should be set according to the granularity of your time series. For example, if your time series contains daily data and you are looking for rolling values over a whole year, you should specify the parameter to window=365. In addition, it is easy to get rolling values for other other metrics, such as the standard deviation (.std()) or variance (.var()).

Instructions

100 XP
  • Compute the 52 weeks rolling mean of co2_levels and assign it to ma.
  • Compute the 52 weeks rolling standard deviation of co2_levels and assign it to mstd.
  • Calculate the upper bound of time series which can defined as the rolling mean + (2 * rolling standard deviation) and assign it to ma[upper]. Similarly, calculate the lower bound as the rolling mean - (2 * rolling standard deviation) and assign it to ma[lower].
  • Plot the line chart of ma.