About visualizations

In the Analysis App window, visualizations are charts and metrics that you can use to visually present, and further analyze, the results output by an analytic. Once you have created a chart, you have the option to export it as an image that can be included in an external report.

The following types of visualizations are available:

Visualization type

Description

Pie chart

Shows categories as a proportion or a percentage of the whole.

Bar chart

Shows comparisons between categories, or between sub-categories.

Line chart

Typically shows or compares trends or changes over time for one or more categories.

Stacked area chart

Typically compares trends or changes over time for two or more categories, and also shows the proportion of the total that each category represents at any given point in time.

Bubble chart

Shows the relation between as many as four different variables, or data dimensions, and is therefore the most flexible chart type. A quantitative data dimension determines the size of the individual data points associated with a category, which allows for comparisons based on the size of data points in addition to their position.

Metrics

Shows statistical information in a visually attractive format for one or more fields in a results table.

The categories referred to in the chart descriptions above are typically non-quantitative data elements such as customers, vendors, departments, branches, merchant category codes, transaction types, statuses, and so on.

The best way to become familiar with the capabilities of the different charts is to use the same small set of data as the basis for each chart type and experiment with different settings. The data set should have at least two fields of each data type – character, numeric, and datetime.

The following sections explain charts in greater detail:

Deciding which chart to use

Which chart you should use often depends on how many data dimensions, or variables, you want to incorporate in the chart. Depending on the type of chart, you can display two, three, or four data dimensions. For example, a simple pie chart or bar chart comparing the total transaction amount for each customer is using two dimensions: amount and customer identifier. A line chart comparing the total transaction amount for each customer for each day of the month of March is using three dimensions: amount, customer identifier, and date (which is actually the dimension of time). Table 1 summarizes the number of data dimensions that each chart type can display.

When you configure charts, each data dimension equates to a field in the table on which the chart is based. Depending on what you want the chart to show, each data dimension can use a different field, or more than one data dimension can use the same field.

Table 1. Number of data dimensions supported by chart type

Chart type

Two data dimensions

Three data dimensions

Four data dimensions

Limitation

Pie chart

Yes

No

No

cannot display datetime data

Bar chart

Yes

Yes

No

cannot display datetime data

Line chart

Yes

Yes

No

Stacked Area chart

Yes

Yes

No

Bubble chart

Yes

Yes

Yes

Chart features

The charts have a number of features that make them a powerful and flexible tool for highlighting and focusing upon different aspects of results data:

Dynamic updating

Charts dynamically update when you do any of the following:

Changes to the data dimensions included in a chart, or the aggregation options, take immediate effect when you click Apply.

Charts also dynamically update if they are contained in an interpretation that has been set to refresh from the most recent version of a results table.

Data point tooltips

When you hover over a data point on a chart a tooltip displays the associated data. If two or more data points have identical chart coordinates, the tooltips are nested.

Interpolated values

There may be occasions when a plotted value on a chart does not exist because of gaps in the data. For example, you could have transaction data for each day in January, but on certain days some customers may have no transactions. In these situations, the gap in the data is interpolated. A line is drawn from the last available value to the next available value. Instead of displaying actual data, the tooltip for the interpolated value displays (interpolated).

Category filtering

Charts in which the displayed data is broken down by category include a legend that lists all the categories, in alphanumeric order, differentiated by color. You can click one or more categories in the legend to remove the categories from the chart. For example, in a chart showing total transaction amount for each customer, you could selectively remove customers to focus the chart on a subset of customers. If only a single category remains, clicking the final category re-adds all categories to the chart.

Optional break field

Charts require at least two data dimensions, which form the X and Y axes of the chart. (The pie chart has no axes.) You have the option of adding one additional data dimension to all the charts, except the pie chart. The bubble chart can include two additional data dimensions. The additional dimension acts like a break field, breaking down the displayed data into multiple categories or sub-categories. For example, if you add transaction type as a dimension to a bar chart displaying the total transaction amount for each customer, the individual customer bars are broken down into clusters of smaller bars representing the amount of each transaction type for each customer.

Aggregation options

All chart types, except the pie chart, support the following aggregation options for the numeric field used for the Y-axis, or for the bubble size component of bubble charts:

When you create a chart, the associated table is automatically classified or summarized as the basis for the aggregation options.

Chart formatting options

Several chart formatting options allow you to control the appearance of the X and Y axes, and modify the size of a chart. For more information, see Creating charts.

Export a chart as an image

You can export any chart as an image (.png file), which you can then include in an external report. For more information, see Creating charts.



(C) 2015 ACL Services Ltd. All Rights Reserved.