Visualize time series

Quiver provides a large number of visualizations on time series. Time series visualizations take one or more time series as input and return a chart.

A time series chart is the most basic time series visualization. Time series charts take one or more time series plots as input and presents the plots on a timeline.

All time series visualizations can be found under the Visualize category in the next actions menu of a time series plot, or in the search bar at the top analysis menu.

• Time series charts
  • Chart card header menu
  • Next actions menu
  • Legend
    • Legend display
    • Plots order
  • Axis configuration
    • Value axis (y-axis) configuration
    • Show/hide value axis (y-axis)
    • Time axis (x-axis) configuration
    • Share an axis configuration with multiple plots
    • Combine axes
  • Zoom control
    • Manual zooming
    • Controlled time (x-axis) zoom settings
  • Time series plots
    • Group plots by data
  • Time and value ranges
  • Date markers
    • Add a marker
    • Delete a marker
  • Ontology Action buttons
  • Plot configuration
    • Rename a plot
    • Interpolation options
    • Chart options
    • Change line display options
  • Move time series plots across time series charts
• Events visualizations
• Scatter plots
• Time-based visualizations
• Area plots

Time series charts

Time series charts are automatically created when you add time series data to your analysis, or when you create new time series plots within the analysis. For example, charts will be created when you transform a time series or create a time series from object sets.

Note that if you add time series data from the search bar and there is already a time series chart in the analysis with an axis of the same unit, it will be added to that existing chart instead of creating a new chart.

From an object selector card, the next actions menu will also offer the object time series property card. This card will add a time series chart configured with the selected series object and its default time series property. You can configure the time series property to instead reference one of the selected object's time series properties or a linked sensor object. Learn more about sensor object types.

Chart card header menu

In addition to card header buttons found in all Quiver cards (Configure, Add to dashboard and More actions), the time series chart header includes extra buttons:

• Add a new time series plot to the chart from the search bar (+Add plot).
• Auto-fit all axes () to automatically adjust the axes display range to a range that contains data.
• Hide/show the legend () to collapse or show the chart legend.
• Download CSV () to download plot data.

Next actions menu

Hover the cursor over a time series plot and the next action menu will display actions which take time series data type as input, such as filtering, visualizations, calculations, transformation and conversion actions.

Hover the cursor over the time series chart itself (any area of the chart that is not the legend area), to have the next action menu display which actions take a time series chart type as input, such as conversion to a transform table.

Legend

The time series chart legend displays all the plots displayed in the chart.

Legend display

To change the legend display:

1. Open the time series chart configuration by selecting the Configure button in the card header ().
2. Select the Display tab.
3. Toggle the relevant settings:
   • Hide legend: Show or hide the legend.
   • Style: Change the style from dropdown to side panel.
   • Inlined/Collapsed: Reduce the legend area by switching to Inlined (when style is set to dropdown) or Collapsed (when style is set to Side) options.

Plots order

Configure the order in which plots are displayed in the legend by dragging and dropping them into the desired order.

Change the legend display to either maximize the chart area used for displaying plot data or maximize the legend display.

Axis configuration

Select an axis label to open its configuration window.

Value axis (y-axis) configuration

From the value axis (y-axis) configuration window, you can:

• Customize the axis label (by default, Quiver uses the unit of an axis as the axis label).
• Manually set the default display range.
• Choose logarithmic scale display.
• Invert the scale (change scale to increase from top to bottom).
• Change the axis format (for example, choosing percentage or currency format, adding a suffix or prefix, choosing scientific notation, displaying negative numbers in parentheses, and more).

Show/hide value axis (y-axis)

Collapse the value axes (y-axis) on one or both sides of the chart to maximize the chart display area by selecting the small triangle icon next to the axis.

To configure whether plots data boundaries (low, center and top) are displayed when value axes are collapsed:

1. Open the time series chart configuration by selecting the Configure button in the card header ().
2. Select the Display tab.
3. Toggle Display y-axes boundaries on collapse on.

Time axis (x-axis) configuration

From the time axis (x-axis) configuration window, you can:

• Customize the axis label (by default, Quiver uses the unit of an axis as the axis label).
• Manually set the default display range.
• Select the auto zoom mode.

Share an axis configuration with multiple plots

Quiver enables you to use the same axis configuration with multiple time series plots. Selecting a Shared axis helps ensure a consistent display across plots of similar data (for example, temperature readings).

To create a shared axis:

1. Open the value axis configuration.
2. Select the Shared Axis button to create a shared axis based on this axis configuration.

When the shared axis button is selected, the value axis label displays the global ID of the shared axis parameter next to it. As displayed in the time series chart configuration panel, the axis is now parameterized to use the shared axis. The dependencies panel also indicates that the time series chart has taken the shared axis as input.

Note that when an axis configuration is parameterized to use a shared axis, the Shared Axis button will not be visible in the value axis configuration window.

To set a plot axis to use an existing shared axis:

1. Open the time series chart configuration by hovering over the plot label in the chart legend and then select the Configure plot button ().

2. Set the plot axis to the desired shared axis parameter.

Combine axes

Axes are combined for plots when their units match. To update a plot's Unit or Unit override configuration, open the time series plot configuration and set the unit value to match the unit of the desired axis.

Zoom control

To facilitate the visualization of your time series data, there are several zoom settings available.

Manual zooming

Hover over an axis for the zoom controls to appear.

Quiver automatically links all time axes across time series charts. As a result, when you pan or zoom one chart's time axes, the zoom range of other time series charts in the canvas will update synchronously.

The value axis (y-axis) controls allow you to:

• Zoom in (+)
• Zoom out (-)
• Automatically adjust the axis display range to the range that contains data by enabling or disabling fit-to-extent.
• Move the axis to the right side of the chart (→).

The time axis (x-axis) controls allow you to:

• Enable streaming (Stream) if streaming services are enabled for this time series
• Pan left (<)
• Pan right (>)
• Zoom in (+)
• Zoom out (-)
• Automatically adjust the axis display range to the range that contains data by enabling or disabling fit-to-extent.

Controlled time (x-axis) zoom settings

Use the auto-zoom settings to change the time axis visible range dynamically. The visible range is controlled by the time axis start and end points. The time axis (x-axis) supports two modes of zoom control:

• Automatic mode: Find the earliest and latest points of plots linked to the axis. This is the default mode for all time series charts.

• Controlled: Allows controlling the endpoints directly or from a time-range source such as a time range parameter. If the Re-enable axis auto zoom on updates toggle is turned on, the auto-zoom mode will switch back to Automatic mode when any of the plots data updates.

Time series plots

The time series chart configuration editor panel allows you to configure basic settings for individual plots on the time series chart. To configure plots within the chart:

1. Open the time series chart configuration by selecting the Configure button in the card header ().
2. Configure the relevant settings:
   • Title: Select the plot name, edit the text, and use the Enter key on the keyboard or click away to save the new text.
   • Remove plot (): Remove the plot from the chart.
   • X and Y Axis: Change the axis to be a shared axis or select Create a new axis to revert back to the chart’s default axis setting. If the chart has been grouped by data, you may need to expand () the plot details to configure this setting.
3. To view the full configuration options for an individual plot, select the Configure plot () button across from the plot’s title.

Group plots by data

By default, plots with the same root object card will be grouped together in the time series chart configuration editor panel and legend. Grouping can be controlled from the time series chart configuration editor panel or the time series chart legend.

In the time series chart configuration editor panel, you can select one of the following options:

• On: Plots will appear under a heading with information and actions related to the root object, ordered alphabetically by global identifier.
• Off: Plots will appear ungrouped and ordered alphabetically by global identifier by default. The plot order can be customized when plots are ungrouped.
• Use default: Plots will be shown according to the Group plots by data setting in the Time Series axes and legends section of the settings panel. This setting controls whether or not plots are initially grouped when creating a new time series chart.

Grouping can be toggled from the legend by selecting Group plots by data () or Ungroup plots ().

When plots are grouped by data, the following options are available in the plot group header:

• Add series from the root object () to the chart. This button opens a menu of time series properties on the root object and sensor objects that are linked to the root object.
• Control with object selection parameter () to update the object references of the plots in the group from a direct object reference to an object selection parameter reference. If there is already an existing parameter with the root object selected, that parameter will be reused. Otherwise, a new parameter will be created with the root object as the selected value. This button is available when the root object is not already an object selection parameter.
• View object selection parameter () to open the parameters panel and highlight the parameter that controls the plots in the group. From there, you can change the parameter’s selected object to update the plots in that group. This button is available when the root object is an object selection parameter.

Time and value ranges

Time and value ranges can be used to highlight and drill down on anomalies in the data, such as specific time periods when an issue was observed or the temperature range when equipment was operating in optimal capacity. Ranges can also be used to enrich the time series data by capturing context on specific time periods or value ranges, such as periods when equipment maintenance took place.

Learn how to configure and use time and value ranges.

Date markers

Date markers are visual symbols (vertical lines) that help identify and distinguish individual data points in the plotted time series. The data points from time series plots that intersect with the date marker are shown in labels on the chart for each plot.

Date markers are controlled by date/time parameters. When adding a marker to a time series chart, a date/time parameter will automatically be added and used as input to the date marker.

Add a marker

To add a marker to a time series chart:

1. Move the cursor over the chart to the desired time and select the chart.
2. Select Save new date marker.

Delete a marker

To delete a marker from a time series chart, hover over the marker in the time series chart and select the trash icon.

Alternatively, open the time series chart editor, navigate to the Markers section and select the trash icon next to the desired marker.

Ontology Action buttons

Action buttons can be added to a time series chart to allow users to write data back to the Ontology. For example, users can create objects, update properties on existing objects, or modify object links.

Learn how to expose Ontology Action buttons directly from the selection menu of a time series chart.

Plot configuration

The time series plot configuration editor panel allows you to configure individual plots on the time series chart, such as:

• Rename plots
• Configure and convert units
• Configure data interpolation
• Override auto zoom settings
• Change the line display options

To open the time series plot configuration, hover over the plot label in the chart legend and then select the configure plot button (). Notice that the configuration button in the chart header controls the settings of the time series chart, rather than the settings of individual time series plots in the chart.

You can also open the time series plot configuration from the analysis contents panel. Simply select a plot and the editor side panel will open on the right-hand side of the application.

Rename a plot

To rename a plot:

1. Select the plot name found at the top of the editor panel.
2. Edit the text.
3. Use the Enter key on the keyboard or click away to save the new text.

Interpolation options

Quiver offers a variety of interpolation strategies to handle missing time series values.

Before: Choose the strategy that Quiver uses to interpolate (infer) the value of a series before the first data point.

• Nearest: Use the value of the nearest point. This option yields a stepped line.
• None: Do not interpolate this series. This option yields discrete points.

Internal: Choose the strategy that Quiver uses to interpolate (infer) the value of a series between data points.

• Linear: Fit a straight line between the two nearest points. This option yields an angled line and is available for numeric time series, but not categorical/enum time series.
• Nearest: Use the value of the nearest point. This option yields a stepped line.
• Next: Use the value of the next point. This option yields a stepped line.
• Previous: Use the value of the previous point. This option yields a stepped line.
• None: Do not interpolate this series. This option yields discrete points.

After: Choose the strategy that Quiver uses to interpolate (infer) the value of a series after the last data point.

• Nearest: Use the value of the nearest point. This option yields a stepped line.
• None: Do not interpolate this series. This option yields discrete points.

Chart options

By default, the zoom level in all time series charts will be automatically extended to fit the combined extent of all plot data sharing the same axis.

Turn off auto-zoom on an individual time series plot to exclude that plot from the auto-extent functionality.

Change line display options

In the Display tab, you can:

• Choose to display a line or points and define their size.
• Configure the line display as solid or dashed.
• Configure the line to appear with or without points.
• Customize plot color.

Move time series plots across time series charts

Time series plots can be moved onto different time series charts by simply dragging and dropping the plot to the chart or into the correct container in the Analysis Contents panel.

Plots can be dragged onto any time series chart on the canvas (when in canvas mode).

Events visualizations

Events are data with start and end time properties. They are represented in Quiver primarily through the event set data type.

There are several ways to visualize and interact with event data within Quiver:

• Events plots overlay events data on other time series plots in a time series chart. This feature can help to contextualize a time series plot or correlate events with observed phenomenon in the time series.

• Some time series plots can take an event set as input, allowing for analysis of series data that occurs during the events, or analysis of the events themselves. For example, the event indicator series takes an event set as input and returns a time series plot indicating the number of events occurring at a given time.

• The events timeline visualizes the events object set over time, segmented into categories if desired. It does not support event sets and only takes in an object set as input.

Scatter plots

There are two methods for visualizing time series data within scatter plots, including:

• Time series scatter plot
• Vega plots: While not limited to time series data, Vega plots can be used to produce custom time series scatter plots. To use time series data as input to a Vega plot, convert the time series plot or time series chart to a transform table using the transformation in the Convert category of the next actions menu.

Time-based visualizations

Quiver enables several additional time-based visualizations, including:

• Time series distribution
• Time series heat grid
• Time series search

Area plots

• Bollinger bands: Plotting Bollinger bands over a rolling time window.
• Time series bounds: Plots the region bounded by two time series.