Charts

Charts in Vesta Online live in a workspace next to the map and the table. You add them from the Visualize hub: pick what you want to see, pick the variables, and Vesta builds the panels. This article adds charts to a copy of the Diabetes and poverty sample, whose US counties dataset has 3,109 counties with diagnosed diabetes and poverty rates. Every step below was run on Vesta Online, and the screenshots come from that run. For the workspace itself, see The workspace.

The twelve steps

1. 1Open the Visualize hubChoose Visualize in the left rail. The page is titled Visualize, Create a workspace, with the choices on the left and a Workspace preview on the right. You need a project you can edit; for a sample, make a copy first.
2. 2Choose what you want to seeFour guided cards cover the common cases. Explore one variable ("Its distribution, spread, and geography.") builds a map, a histogram, a box plot and a table for the variable you pick. Compare two variables ("Maps side by side plus their relationship.") builds two maps, a scatter plot and a correlation matrix for variables A and B. Explore many variables ("Correlations and patterns across several at once.") is the third guided card. Show an analysis result ("Maps, plots, and summaries from a finished analysis.") builds panels from an analysis you have already run. Here, choose Explore one variable and pick Diagnosed diabetes (% of adults).
3. 3Create the workspaceChoose Create workspace. Vesta adds a new workspace tab with the four panels from the preview. Each chart draws from your data, and a Statistics panel beside the grid lists the count, minimum, maximum, mean, standard deviation and more for the variable.
4. 4Compare two variablesGo back to Visualize, choose Compare two variables, set A to Diagnosed diabetes (% of adults) and B to Poverty rate (% of people), and choose Create workspace. The scatter plot draws one point per county, and the Statistics panel adds the correlation (r) and the rank correlation (ρ) of the pair.
5. 5Open the chart settingsChoose the gear in a chart panel's header. The Visualization settings dialog opens over the workspace. A map has its own Map settings instead.
6. 6Change the titles and labelsThe dialog has four boxes: Panel title, then under Appearance Chart title, X Axis Label and Y Axis Label. Type a new value in any of them. On the scatter plot used here, the walk typed all four values one after another, and the new title and labels showed in the panel without an apply step; the dialog is marked Changes apply immediately. The dialog also has Reset all.
7. 7Close the settingsChoose Done, or press Escape. The chart keeps its new title and axis labels, and the workspace saves the change: after a reload the scatter plot still showed all three.
8. 8Change a histogram's titles and labelsOpen the Explore one variable workspace again and choose the gear in the histogram's header. Its dialog has the same four boxes and a Bins section with Number of bins, here 10. Type a Chart title, an X Axis Label and a Y Axis Label, then choose Done. The histogram draws the new chart title and both axis labels, and they were still there after a reload.
9. 9Build a workspace by handOn the hub, choose the Manual card ("Build the workspace yourself — layout and every panel."). Pick a layout with the four buttons: One panel, Two side by side, One large, two stacked, or Four panels. Choose a slot's header to open the catalog of views.
10. 10Pick a chart type and its variables for each panelThe catalog groups views as Maps, Charts, Tables and Results. The Charts group lists Histogram, Cumulative Distribution, Box Plot, Scatter Plot, Heatmap, 3D Scatter Plot, Correlation Matrix and Time Plot. After you choose a type, pick its variables at the bottom of the slot: one variable for a histogram, cumulative distribution or box plot, two for a heatmap or scatter plot, three for a 3D scatter plot. This walk-through fills four slots with a cumulative distribution, a heatmap, a 3D scatter plot and a histogram, then chooses Create workspace.
11. 11Show an analysis resultChoose Show an analysis result and pick the analysis output from the list. The sample ships with a Local Cluster Analysis, so the list includes Moran Scatterplot. Choose it and Create workspace; the Moran scatterplot opens in its own workspace, colored by cluster class.
12. 12Add a chart to a workspace you haveIn a workspace, choose Add panel under the grid. The hub opens with Create new and Adjust current, and its main button reads Update workspace. This walk-through opened the hub from Add panel and did not build a panel from it.

What you should see

The hub

Visualize

The hub opens on Compare two variables, with a preview of the four panels it will build.

Explore one variable

Explore one variable

The guided card fills the preview: a map, a histogram, a box plot and a table for one variable.

The workspace it creates

Workspace

In the walk-through, the histogram groups the 3,109 counties into 10 bins. The Statistics panel gives count, minimum, maximum, mean, standard deviation, skewness, kurtosis, median and interquartile range.

Compare two variables

Scatter plot and correlation matrix

Diabetes against poverty, one point per county in the walk-through. The correlation matrix has a Zoom list.

The settings dialog

Visualization settings

A live dialog: Panel title, and under Appearance the chart title and the two axis labels.

New titles and labels

Values typed in the four boxes.

The result

The scatter plot panel, chart title and both axis labels show the new text.

The histogram, edited

The histogram with the chart title Diabetes across US counties, the X axis label Share of adults with diabetes (%) and the Y axis label Number of counties.

The manual catalog

Catalog

Choosing a slot's header replaces the guided cards with the catalog.

Four panels chosen

Each slot has a chart type at the top and its variables at the bottom.

The manual workspace

The 3D scatter plot: drag to rotate, scroll to zoom, double-click to reset.

The Moran scatterplot

Show an analysis result

The result list holds the sample's finished analyses.

Moran scatterplot

Each county the analysis tested is a point; the analysis tested 3,105. High-High points sit in the upper right, Low-Low in the lower left.

Add panel

Add panel

Opened from a workspace, the hub can adjust the current one.

Reading the charts

Chart	Variables you pick	Where it appeared in the walk
Histogram	One	Explore one variable; the manual workspace
Cumulative Distribution	One	The manual workspace
Box Plot	One	Explore one variable
Scatter Plot	Two	Compare two variables
Heatmap	Two	The manual workspace
3D Scatter Plot	Three	The manual workspace
Correlation Matrix	Not picked in the walk	Compare two variables
Time Plot	Not walked	Not walked

The Moran scatterplot is the scatter plot a Local Cluster Analysis saves. Its chart title starts with “Moran Scatterplot: “ followed by the variable name, and the points are colored by cluster class. To make one, run the analysis as described in Local cluster analysis and choose it under Show an analysis result.

NoteThe settings dialog holds a panel title and, on the scatter plot and the histogram walked here, a chart title and two axis labels. The histogram's dialog also has a Bins section with Number of bins, which this walk read (10) but did not change.

NoteNot walked: Explore many variables; Reset all; closing the dialog with Escape; a panel's menu (Export, covered in Exporting; Rename panel…, Move left, Move right and Remove panel); the map's own Map settings; building a panel from Add panel; the Time Plot, Contingency Table, ROC Curve and the other result charts. From the product's code, not from this walk: the box plot has no legend, no tooltip, no zoom, and the Time Plot shows at most 200 lines at once.

NoteThis walk-through added a histogram, box plot, scatter plot, correlation matrix, cumulative distribution, heatmap and 3D scatter plot, and a Moran scatterplot from a saved result, so the screenshots show those.

Where next

• How panels, tabs and layouts work: The workspace.
• Make the result behind the Moran scatterplot: Local cluster analysis.
• Work on a map next to the chart: Maps.