The original Combo chart displays multiple data series on up to two separate axes as line or vertical bar charts.

The optional multi-series capability of the Combo chart can break out values by additional dimensions using up to two Y axes.

You can also choose multiple data sources and display data together on the same chart. There is no limit to the number of data sources you can display on a single chart, up to the capabilities of your chosen browser. You can use the same data source more than once, and choose a different set of columns to display.

Zoom and Pan

You can scroll the mouse wheel or track pad to zoom in and out on the data in your combo chart. This is equivalent to manually brushing over the range chart to choose a zoomed in area.

If you scroll while holding the shift key, the range box pans left and right. This lets you zoom in on a section (normal scrolling) then move left and right to find the exact data you want to see.

You can zoom in or zoom out regardless of whether the range chart is currently displayed. Turning the range chart on or off does not clear a range brush filter.

If you select an area of the focus chart by brushing, then change the range chart in a way that makes the selection on the focus chart no longer visible, the filter on the focus chart is cleared.

A Reset button appears on the lower right as you mouseover a chart with a range filter in effect. Clicking the Range button clears out the range filter only — a brush applied to the focus chart remains.

Combo Chart Examples

Create a new Combo chart. Choose a Data Source. This example uses the flights sample database.

Categorize on the Dimension plane_issue_date, the date the plane was first acquired. To see how on time performance compares to the aircraft acquisitions over time, set the measure for Y Axis 1 to depdelay, and the measure for Y Axis 2 to arrdelay. These measures are within the same range, and combine well on the chart.

You might also want to see how frequently flights are cancelled, based on the age of the airplane. Add the cancelled field as Y Axis 3.

The difference in values makes it difficult to see how the trend of cancelled flights compares to flight delays. Click the line next to the cancelled item in the COLOR PALETTE list and choose Secondary Axis.

You can make the distinction clearer still by changing the secondary axis to a vertical bar chart.

You can also change the color and style of the lines in the chart to make them more distinct.

When you use a numerical dimension, you have the option of displaying the chart in Percentage View. The pop-up box displays not only the values for each item on the X axis, but also the percentage of the total values for that group.

Multisource Combo Chart Example

This example uses a dataset of crimes committed in San Francisco and another dataset of crimes committed in Seattle. We can compare them both on the same axis. Create a new combo chart, and select sfcrime as the Data Source. Choose DateEvent as the X Axis and #Records as the Y Axis.

Click Add Another Data Source and choose crimes_seattle. Set the X Axis to At_Scene_Time and the Y Axis to #Records.

The datasets do not overlap completely. You can use the Range Chart to select a range for fair comparison. In this example, 01-01-2014 to 01-01-2015.

Edit the axis labels and title for clarity. Set the X-axis minimum to 0 and maximum to 900 to match the X-axis for Seattle. Click Apply. Your chart is ready for display on your dashboard.

Charts in this section are deprecated; these charts have been replaced by the improved Combo chart type (formerly known as “New Combo”). On dashboards with older charts, you can open the menu at the top of the dashboard or at the top of each older chart to upgrade the chart. These chart types will be removed in a future release of HEAVY.AI. Creating new charts with these types is discouraged, and upgrading each of the older charts may be required prior to upgrade. If you wish to create additional charts using these deprecated chart types, you are currently able to do so in by setting the ui/hide_deprecated_chart_types to false, which will allow these to appear in the listing of chart type options. Users are encouraged to contact support () with any questions or concerns regarding deprecation and removal of these chart types.

Horizontal Bar charts display values for multiple dimensions, with two measures displayed as the width and color of the bar for each dimension group.

Bar charts are useful for showing relative values based on categories of information, particularly when the category label is long.

Sort By

You can sort by any chosen dimension or measure in ascending or descending order.

# Of Groups

Display up to 100 groups of records. You can enter a value or use the slider to visually set the number of groups.

Null Dimensions

Choose whether to show or hide Null values for your chosen dimension.

Color Palette

You can use a custom palette to visually group values in your chart. By default, data points are colored arbitrarily with the same solid color. You can change the color for all columns, or apply colors to individual Dimension values.

If you set the Color measure, you can choose a gradient to visually express relative quantitative values.

Custom Measure Formatting

You can also use custom measure formats for the values in your chart. See Customizing Measure and Date Formats.

Bar Chart Examples

Create a new Bar chart. Choose a Data Source. This example uses the official database of UFO sightings

Bar charts require a minimum of one dimension and one measure. For example, this chart categorizes on the Dimension of shape and uses the count of each shape as the Measure.

You can add any number of dimensions to your Bar chart. For example, you can combine the shape with the state where they were seen.

When you use two measures, the first measure (shape) is indicated by the width of the bar, while the second measure (Duration_Seconds) is indicated by the color of the bar. Reversing the Color Palette makes the labels easier to read, in this example.

You can use a custom palette to color by dimension and visually group values in your bar chart. For example, this chart assigns colors to similar shapes (for example, rounded shapes such as disk, circle, sphere, and oval are all colored orange).

The Histogram displays the distribution of data across a continuous (typically time-based) variable, by aggregating the data into bins of a fixed size. Vertical bars show the count of data within each bin.

Histograms can count occurrences of data other than the binned dimension by assigning a series.

Use a Histogram to understand the distribution of your data, and to see areas of unusually high or low density, which would be masked by an aggregate such as Average.

Once you choose your measure and dimension, you can edit the labels for the X and Y axes. Click the label and enter your custom text.

Show Range Chart

You can change the range of values in your chart by "brushing" over the range chart, or by entering fixed start and end values on the X Axis.

Color Palette

You can use a custom palette to choose the bar color in your chart. By default, data points are colored light blue. You can choose one of 8 colors.

If you set the Color dimension, you can assign colors to individual values, or you can assign the same color to several dimension values to visually group them in the chart.

Custom Measure Formatting

You can also use custom measure formats for the values in your chart. See Customizing Measure and Date Formats.

Histogram Examples

Choose a Data Source. This example uses the official database of UFO sightings

This chart categorizes UFO sightings on the dimension of Sighting_Time and uses the number of sightings as the measure.

Adding the Color dimension of shape visually shows the relative number of sighting types within each time-based bin.

The records go back to 1905, with very few recorded sightings. Selecting a more recent 10-year block presents information in a more digestible form. For an ad hoc report, you can brush across values in the Range Chart to filter a smaller number of bins, or you can enter start and end dates on the x axis for precise values.

When you set the x axis to a date/time value, you have the option of setting the BIN size to month, quarter, or year. Setting the BIN size to 1y lets you compare the aggregate values at a glance.

Once you set the boundaries for the x axis, you can turn off the SHOW RANGE CHART setting. You can also reduce the number of series displayed by removing their entries on the COLOR PALETTE.

You can change the labels on the axes and the chart title to reflect the current state of the information.

Stacked bar charts display values for one or two dimensions, one of which must be defined on the x-axis. Measures are displayed as the height and color of the bar for each dimension group. With stacked bar charts, you can display multiple measure/dimension combinations on each bar. This makes it easy to visually compare measures for the defined dimensions, as well as see the cumulative value for the defined measure.

Stacked Bar Chart Examples

The following examples use the flights_2008_7M sample database.

Stacked Bar charts use two dimensions (X axis and color) and one measure.

1. Create a new Stacked Bar chart.

2. Choose flights as the Data Source.

3. For Dimensions, select destination state for X Axis and origin state for Color. To remove the null dimension on the X Axis, toggle Null Dimensions to off. Note:

   • If you omit the Color dimension, you create a single-color vertical bar chart.

   • When you use the Color dimension, Immerse selects the top five categories for the dimension and measure. When the All others toggle is on, all categories not in the top five are shown as Other.

4. For Measures, select the sum of airtime.

5. Sort by descending airtime, and define the colors for origin states. Note: You can enable and disable All Others Color Palette, but you cannot add or remove categories.

This creates a chart like the following, showing the cumulative airtime for flights, based on origin state and destination state. The origin state airtimes are stacked on the same bar, making it easy to see total airtime by destination as well as total destination airtime by individual origin.

Now, adjust the chart dimensions and measures to see information about cancelled flights for particular origination cities for individual airlines.

1) For Dimensions, select origin city for X Axis and select carrier name for Color.

2) For Measures, select the sum of cancelled flights.

This creates a chart that shows the total cancelled flights for particular origin cities, broken out by air carrier.

If you prefer, you can set the Percentage View toggle to compare relative values within a dimension and side-by-side.

You can reduce the number of cities on the X Axis by sliding the # of Groups down (15 in the next example), and show the value for all other airlines by toggling All Others to On in the Color Palette:

In some cases, if you disable All Others from the color menu, the # of Groups does not match what is shown on the x axis. For example, in the following chart, the number of X axis items for dest_state is five, but the # of Groups is set to 6.

This occurs because the query that is returned has "undefined" for "key1" values, and these values are filtered out by Immerse. When the All Others toggle is enabled, these values are referred to as "other" instead and are grouped together, making the number of values on the X axis match the number defined in # of Groups.

The Line chart represents a series of data as a line or multiple lines, plotted across time or another numerical dimension.

The optional multi-series capability of the Line chart can break out values by an additional dimension.

Line Chart Examples

Create a new Line chart. Choose a Data Source. This example uses the official database of UFO sightings

Categorize UFO sightings on the Dimension sighting_time and set the Measure to the number of records (the number of sightings).

Adding the Dimension shape displays a separate line for each sighting type across time.

The records go back to 1905, with very few recorded sightings. Selecting a more recent 10-year block presents information in a more digestible form. For an ad hoc report, you can brush across values in the Range Chart to filter a smaller number of data points, or you can enter start and end dates on the x axis for precise values.

When you set the x axis to a date/time value, you have the option of setting the intervals to month, quarter, or year. Setting the BIN to 1y can make the values easier to discern.

You can change from a line chart to an area chart, adding more contrast to make the differences between the values more apparent.

You can also reduce the number of series displayed by removing their entries from the COLOR PALETTE.

After you reduce the scope of the chart, you can change the labels on the axes and the chart title to reflect the current state of the information.