CanvasXpress Visualization Library

The following chart types have layouts that depend on chart dimensions and may vary across different browsers and platforms:

• Treemaps: Sample locations are area-based; layout depends on chart size.
• Tag Clouds: Layout is dynamic and depends on chart dimensions.
• Layout Charts: The order of elements may vary.
• 3D Scatterplots: The superimposition/stacking order of points and 3D overlays (like ellipses in scatter3d2) may differ.
• 3D Scatterplot Transpose: Transposing 3D plots will result in only a few points; this is intended behavior.

If you set resizableX or resizableY to FALSE, these restrictions apply to the 'maximize' button in the menu bar as well:

• If resizableX is FALSE: Maximizing the plot will expand the height but the width will remain fixed.
• If both resizableX and resizableY are FALSE: The plot will not resize when maximized and may move to the top-left corner of the container.

You can install the stable version of canvasXpress from CRAN, or install the latest development version directly from GitHub using devtools.

# Install latest version from GitHub
devtools::install_github('neuhausi/canvasXpress')

The canvasXpress package includes a function cxShinyExample() to run interactive Shiny applications demonstrating advanced features like mouse events, zooming, and broadcasting.

• Use cxShinyExample() to list available example names.
• Use cxShinyExample(example = "name") to run a specific example.

# List example names
cxShinyExample()

# Run an interactive shiny example
cxShinyExample(example = "example1")

To learn more about getting started or to see additional examples, you can access the built-in R vignettes.

• vignette(package = "canvasXpress") lists all available vignettes.
• vignette("getting_started", package = "canvasXpress") opens the getting started guide.
• vignette("additional_examples", package = "canvasXpress") opens the additional examples guide.

# List all package vignettes
vignette(package = "canvasXpress")

# View a specific vignette
vignette("getting_started", package = "canvasXpress")
vignette("additional_examples", package = "canvasXpress")

The sankeyType property determines how the strata (columns) are scaled in a Sankey/alluvial diagram:

• "normal": One shared scale across the whole diagram. Columns with fewer classes appear shorter (default).
• "proportional": Each stratum column is scaled independently to fill the full plot height. Node heights remain proportional to their values, and flow ribbons fill the node boxes proportionally at both ends.
• "equal": Every stratum (node) in a column gets the same height, partitioning the column evenly. Ribbons fill the equal node boxes proportionally.
• "aligned": Keeps normal node heights but spreads the strata of each column evenly over the full plot height (this overrides sankeyNodePadding).

{
  graphType: 'Sankey',
  sankeyType: 'proportional'
}

For very large charts, you can enable yieldingRender to keep the page responsive.

When yieldingRender is set to true, the 'cold render' yields to the browser after data ingestion. It paints recorded objects in short requestAnimationFrame slices instead of one long task. This makes rendering asynchronous.

To know when rendering is complete, monitor the CanvasXpress.ready event or the afterRendering flag.

Default Behavior

• yieldingRender: false: Rendering is fully synchronous (ensures identical pixels).

Control how data points appear when selected using selectionMode:

• "highlight": Recolors the selection (default).
• "ghost": Fades the non-selected marks.
• "focus": Recolors non-selected marks to grey and keeps the selection in color (storytelling focus).
• "name": Labels the selection (requires selectionMetadata to be set and Scatter plot).
• "filter": Keeps only the selected points.

Selection Properties

• selectedDataPoints: Array of variable/sample names to initially select (default: []).
• selectionMetadata: Metadata (from data.z) to use for labeling when selectionMode is "name".
• selectionColor: Color for selected points when selectionMode is "highlight" (deprecated; use highlightColor).

CanvasXpress allows for data transformation during the loading process using aggregations and calculatedFields.

Aggregations

Used for Tableau-style materialized group-by operations. Each entry follows the pattern: { by: [annotationNames], defaultFn, overrides: { variable: fn }, sep? }. Supported functions (fn): mean, median, sum, min, max, sd, count, first.

Calculated Fields

Allows creating new columns based on formulas. Each entry is { name, target, formula }.

• target: Can be variable, sampleAnnotation, or variableAnnotation.
• formula: A safe expression (e.g., log2(A/B), (V1 - mean(V1)) / sd(V1)).

These are applied after calculatedFields and are serialized so they persist with the chart.