The Parallel Coordinates visual is a line chart which plots values over multiple dimensions rather than over a continuous series. Each data point has its own vertical axis and can therefore plot data which have different scales and formats. The parallel coordinates chart can help to highlight patterns in datasets that may not otherwise be clear in traditional time series charts.
In contrast to a time series chart where performance is measured by changes over the whole series, the difference in the parallel coordinates chart is that the value lies in identifying patterns. For example, using the chart above as an example, the towns where a high percentage of the population is over 60 also have a low percentage of people with a tertiary degree. The towns with a higher average income also generally have a younger population. Outliers can also be easier to identify. The parallel coordinates chart makes it easy to see that Town J has a higher than average income and a high population of people over 60. These sort of insights would be incredibly hard to identify with a time series graph.
The Parallel coordinates chart requires the legend field to be populated and one or more measures provided in the values field. Only numeric attributes can be used in the Values field. Non-numeric attributes will not display. Each measure used in the values field will display on its own vertical axis.
Because each measure has its own axis, the measures can be different scales and formats.
Note: Measures formatted as a decimal number where numbers are
The order of the axis will depend on the order the attributes were added to the Value field, so it is important to select the order required first, otherwise you will have to remove and re-add the fields in the required order.
Clicking on a line in the chart will highlight the selected line and grey out the remaining lines.
Hovering the mouse over a line will display a tooltip with the value of each data point along that line.
Limitation: Always use measures on the chart to enable correct scaling. Columns with numeric attributes will not always display correctly.