Visual Inspector - additional charts
The Horiz Histogram is the same as the Histogram, but bins along the Y axis, depicting the aggregation along the X axis. The Horiz Histogram is always binned.
The Box Plot depicts (typically) a category along the X axis and a numeric column along the Y axis. For each category, the average, quartiles, and outliers are depicted.
Quantile shows the quantile of each value of a single column. The value is specified on the Y axis, and the quantile on the X axis.
A Path plot is similar to a Line plot, but the line is not drawn left to right. Instead, the line is drawn in the order of the data. The two columns specify the points of the line, and the points are simply traced with a line.
A Map plot is a Scatter over an underlying geographic map. The X and Y axes should be longitude
180) and latitude (values from
To use the Map plot type, a Google Maps key must be provided. Once an API key is obtained, it can be used by setting an environment variable. Refreshing the Inspector once this has been done will enable the Map chart.
`AX_GOOGLE_MAPS_KEY setenv "mykey"
Both of the following environment variables must be set to use the MapBox layer.
`AX_MAPBOX_KEY setenv "mykey" `AX_MAPBOX_STYLE setenv "user/styleid"
The MapBox style ID is found as part of the URL for any MapBox style in
studio.mapbox.com. Given a style
user/id segment is used in the environment variable above.
A Plot Matrix renders every column of the table. If n columns exist, a n*n chart will be displayed. The diagonal of the chart is a single Histogram for each column. Every cell not in the diagonal is a pairing of the two columns along the diagonal that define its coordinate. Note, it is a good idea to use this chart only on tables with few columns since each additional column requires more render time, and produces a smaller chart in each cell of the matrix.
The Visual Inspector supports the visualization of Networks (directed graphs) through a force-directed layout.
There are two ways for data to be formatted so that the Directed Graph visualization can determine the parent-child relationships. First, a parent and child column can simply be supplied. This would look like:
Parent | Child -------|------- A | B A | C C | B
Alternatively, each entry in the child column could be a list of all children for the given parent. This allows for more compact tables. For example:
Parent | Child -------|------- A | B C C | B