From the course: Power BI Data Dashboards

Analytics options: Reference, trend, and forecast lines - Power BI Tutorial

From the course: Power BI Data Dashboards

Start my 1-month free trial Buy for my team

Analytics options: Reference, trend, and forecast lines

“

- [Instructor] The analytic options for a Power BI visual enable us to add elements like reference lines, trend lines, and forecast lines. The available options depend on the visual that we're looking to add these analytics to. Let's take a line chart for the US population over time. First though, let's make sure we start the y-axis is zero, so we get an accurate perspective on the magnitude of the population. We can then add reference lines to it. Let's say that we add a reference line to the y-axis to visually denote where the US population passed 300 million people. We'll access it through the analytics pane, then choose to add a y-axis constant line to the visual, which we'll set to 300 million. We can then add a data label to this line as well. We can add multiple lines to the visual from the analytics options. We can choose to add a trend line to show the overall direction over time. This line uses linear regression to fit a line amongst the data points. The US population trends up over time but adding a trend line enables us to see what the smooth trend looks like. Let's make this trend line orange so it stands out against the background and also against the blue line of the actual data points, in this case, the population over time. If we change the selected state, we see the trend line changes accordingly as well but the y-axis constant line stays the same. Another option we can add within the analytics pane is forecasting. The forecasting model uses past data points to make predictions about future ones. Forecasting Power BI uses time series exponential smoothing. Let's forecast out the next five years of our data. We can adjust the inputs for the forecast within the forecasting menu. The solid line indicates the most likely forecast while the shaded area around it illustrates the confidence interval. By default, Power BI uses a 95% confidence interval but we can change it to a wider or narrower confidence interval band to predict our future data points. Forecasts in general can be difficult to make, and even though this analysis is useful, we would still want to supplement it with other information analysis before making a final recommendation for forecasting. If you want to check out more on Power BI's forecasting capabilities, check out the Power BI: Integrating AI and Machine Learning course in the library.

Contents