Chart overview

Charts let you visualise the data from your Data Panel. Many types of charts are available.

Between the standard Chart Types and the Custom Vega-Lite charts available, there is a vast amount of customisability for visualising your data. You may want to play with multiple options to see what works best for the story you are trying to tell.

Most charts use X-axis, Y-axis, Colour, and Facet, and can accomodate formatted labels.

Creating a Chart

Regardless of the type of chart, you will create a new chart panel using the icon. Place the chart panel wherever you want it on your dashboard.

Click on the Chart Type lozenge to start creating the new chart

Chart Type

The chart type should reflect the data types being visualised.

All charts have a continuous secondary axis, representing an aggregate calculation:

Area chart: continuous variables for both axes, with interpolation options

Line chart: continuous variables for both axes, with interpolation options

Bar chart: one discrete axis and one continuous axis

Circle, Point, Tick charts: Functionally these are the same, with different markers. The primary axis can be continuous or discrete.

Custom: Anything possible using Vega Lite specifications can be created here.

Interpolation

Interpolation is a way to approximate values between data points

If you have selected Area or Line charts you may want to go back to the chart type menu and select a method to interpolate.

The interpolate options are

****Linear (piecewise linear segments, as in a polyline)
Step (i.e. piecewise constant, or nearest-neighbor; alternate between horizontal and vertical segments)
Basis (a B-spline, with control point duplication on the ends)
Cardinal (a Cardinal spline, with control point duplication on the ends)
Monotone (cubic interpolation that preserves monotonicity in y-axis)

X-axis and Y-axis

By default, the X-axis pops up in chart creation as the primary axis; however, charts can instead be created with the Y-axis as the primary axis by leaving the X-axis variables undefined until after the Y-axis is defined. For example, a horizontal bar chart is possible by first defining the Y-axis as the discrete variable and then selecting the aggregation on the X-axis.

TIP: To clear the axis data (e.g. to transpose your axes) simply click the "x" in the Axis Lozenge.

Colour Column

Choose the series to be coloured (which variables are charted against each other for comparison).

TIP: To remove colour option after adding it, simply click the "x" in the Colour Lozenge.

NOTE: To change the actual colours, you need to use the overall dashboard controls

Selecting a colour column requires choosing the column's characteristics and how the chart should be rendered:

Data Type (Quantitative, Temporal, Ordinal, Nominal) affects the default colour palette (categorical or gradient). Change this manually from Auto to one of the other options to see how the visualisation changes.
Sort (ascending or descending, according to your secondary axis values)
Stacking (Stacked, Normalised stacked, Row, or Overlapping)

Stacked view & Normalised stacked view

Values as discrete or a percentage of the total

Row view & Overlapping row view

With Row view, all secondary axes are scaled to fit the largest values. With Overlapping Row View, the secondar axis is not necessarily scaled to fit, and the largest values overlap the other colours.

Overlapping row view: Ascending & Descending

Changing from Ascending to Descending (or vice versa) may help to prevent some data from being hidden.

Facet Column

While the Colour Column plots values separately for each Colour Column value, they all use the same primary axis.

The Facet Column splits them into separate charts. Setting the number of columns and rows changes the display layout of the charts. Setting one row per facet makes a very similar chart to selecting the "Row view" colour options.

TIP: To remove facet option after adding it, simply click the "x" in the Facet Lozenge.

See the same data below, with variations in the colour and facet column options.

Changing the number of columns and rows in the facet options will only affect the compactness of the display, not which data gets shown. It is essentially a way to tell the software how many charts you want to see on your screen at one time. If your data splits into ten facets (ten charts) and you set 1 column & 1 row, you will need to scroll to see each of the charts. This can be useful to manipulate if you want to allow more or less detailed viewing of each chart.

Label angles & size

Changing the label angles and size can be critical to legibility and fast interpretation. Labels can be vertical, horizontal, or diagonal. Changing label angles helps legibility and often makes better use of space.

Label sizes change the number of pixels used by the labels, so must sometimes be increased to show longer labels.

Compare the following default labels (left) and edited (right)