General Coding Patterns

General Coding Patterns

The arguments each chart type accepts are slightly different reflecting the nature of each plot. However, there are some common elements found across many plot types:

Filters

Filters allow users to subset the data that is used to display the chart. The logic is that a datapoint is included if it satisfies ALL of the various filter conditions.

Filters can be specified with either a Dict or a Vector:

• Vector of Symbols: filters = [:product_line, :country] creates multi-select dropdown filters with all unique values selected by default. So if you have [:product_line, :country] then you will end up with two dropdown boxes allowing you to subset the data by country and by the product_line.
• Dict: filters = Dict(:country => ["Australia", "United Kingdom"], :product_line => nothing) where each key is a column name and each value specifies the default selected values. This is a bit more flexible than the above as you can specify what filters are applied by default. In this case you will end up with the same filters being possible as in the above case but by default when a user first opens a page there will be filters applied for Australia and the UK.

Color Columns

color_cols are the options that affect the grouping and color of objects in the chart. For instance in a KernelDensity plot, different colors can be used to show the distribution of different groups. In a ScatterPlot, different colors can be used to show points from different groups.

Facet Columns

facet_cols specify which columns can be used for faceting (creating small aligned multiples of the same plot).

Note that if only one facet_col is specified there will be one dropdown list for applying this faceting or not. If two or more are provided there will be two dropdown lists for the x and y dimensions of the facet_grid.

• facet_cols: Defines which columns are available for faceting. Can be a single Symbol or Vector of Symbols.
• default_facet_cols: Specifies which facet columns should be used by default:
  • If nothing, no faceting will appear initially (but users can enable it if facet_cols is provided)
  • If specified, the chart will start with faceting enabled using those columns

Value Columns

Many chart types accept value_cols which specify the columns containing the data to visualize. For plots that can display multiple value columns (like DistPlot or KernelDensity), you can provide a vector of column names and use dropdown controls to switch between them.

x, y, z Columns

Some chart types allow you to specify several columns that are used for x, y and z. In these cases you can swap them on the fly.

Interactive Controls

Several chart types have their own interactive controls that are quite specific to them. Some examples are:

• The DistPlot has a button to turn on or off the rugplot;
• The ScatterPlot has buttons to turn on or off the density shading
• The KernelDensity has a slider beneath it to change the kernel bandwidth
• ScatterSurface3D has a button to change the regularisation of the surface from L2 to L1

and there are many more. For some details on these see the specific section on the plot in this documentation.