Skip to main content

Identify Points

The Identify Points options in Add to Plot allow you to label, colour, and highlight specific points of interest in your plots. This is useful for investigating outliers, highlighting specific observations, or adding context to your visualisations.

The identify panel has two parts: choosing how to label identified points, and choosing how to select them.

How to label points​

  • Text labels: Select a variable from the dataset to label identified points. The label text appears next to each point — typically a name, ID, or category. When no label variable is selected, row numbers are shown instead.
  • Colour points: Fill identified points with a chosen colour to make them stand out from the rest of the data (default: red). Identified points are drawn as solid filled symbols for emphasis.
  • With the same level of: Highlight all observations that share the same level of a grouping variable as the identified points. For example, in the Gapminder dataset you can identify a country in one year and track all observations of that country across time. The originally identified point is highlighted for reference.

How to select points​

Click to locate​

You can click directly on the plot to identify the nearest point. The Click to Locate button puts the plot into an interactive mode — click on any point, and iNZight finds the closest observation and labels it. You can click multiple times to build up a set of identified points, and use the Click to Remove button to deselect individual points by clicking on them.

Limitations

Click-to-locate is not available when:

  • Multiple plots are displayed (e.g., when using subsetting variables with _MULTI)
  • For dot plots, when Variable 2 is a categorical variable

By value of a variable​

Select a variable and a specific value to highlight all observations matching that value. For variables with a small number of unique values (or a factor), a slider is provided to quickly cycle through levels. For numeric variables or when you want to select multiple levels, you can click the Select levels button. This is useful for exploring how specific groups are distributed across the plot.

By extreme values​

Select the most extreme points automatically. For scatter plots, this uses Mahalanobis distance from the centre of the data, identifying points that are most unusual in terms of both their x and y values considered together. For dot plots, you can separately control how many points to identify from the lower and upper ends of the distribution using the N Lower and N Upper sliders.

Once identified, you can save the selection using the Save these points button if you want to retain the same observations across multiple plots.

Applicable plot types​

Point identification is available for scatter plots and dot plots, as these plot types display individual data points that can be labelled and highlighted.