Time-series Table

Overview

Time-series Table charts combine a tabular presentation of data with time series columns & charts that convey data point snapshots and progression over time. With a Time-series Table, you can easily detect anomalies in a specific cohort while also seeing trends across all cohorts.

Reference Content

The following articles may be useful resources as you build your chart:

• Creating a Chart: A walkthrough of the overall process of chart selection and creation.
• Using Preset Chart Builder: How to work with the Chart Builder page, the primary interface used when creating a chart.

Creating a Time-series Table Chart

Preset currently has a few options for table charts in the visualization picker:

• Table Chart
• Time-series Table Chart
• Pivot Table

Look for the Time-series Table chart type—or use the Search feature to filter by the word “Time”—and then select.

To create a Time-series Table, you need to define in the Chart Builder:

• the Time Column you want use for any time-series visualizations
• the Metric(s) that is being plotted over time
• the Dimension(s) you want to use for to define new rows of the table

We will walk through an example of finding average ride times for different types of Citi bikes from the public Citi Bike System Data.

	Time Column Like other time-series charts, the time column is the single column used for all time-series visualizations on the table. In this example, we want the ride's start time as the X Axis value for our time-series and our time column.
	Metrics Metric(s) is the core value being plotted across time for each dimension of the table. In this example, we are creating a custom metric for the average ride time. Dimensions Dimensions can be any categorical value that is used to segment the data further. In this example, we are looking to understand if the average ride time and ride patterns are different for different bike types.
	Time Series Columns After the above fields of Time Column, Metrics, and Dimensions are defined, we need to further specify how the metrics are displayed. There are many built-options, including Sparkline, Time comparison, Contribution and Period average.The subsequent rows added show up as additional columns on our table. We will choose Sparkline for the first column of Time-series and Time comparison for the another column. For Time comparison value, we know our data has daily granularity, so a Time lag of 28 will show the Actual value from 28 days ago.

Different Column Configurations

There are 4 different types of Column Configurations available for the table columns:

• Time comparison: Produces data points as a result of a direct comparison between time points (e.g., percentile change week over week - WoW).
  • Actual Value: Shows the metric value without any time comparison.
  • Difference: Shows the difference between the last data point vs. the data point of Time Lag period granularity.
  • Percentage: Shows the percentage that represents the last data point vs. the data point of Time Lag Period Granularity.
  • Percentage Change: Shows the percentile change based on the first and last data points within the Time Lag period.
• Contribution: Shows the contributory percentage of the row metric relative to the total of all rows.
• Sparkline: Shows a line chart of metric values for the time range specified.
• Period average: Shows the average of the numbers of periods specified in the Time Lag.

Each configuration provides a different set of additional settings:

• Label: Column title name
• Tooltip: Hover tooltip information for column title
• Type: Type of configuration (comparison, value difference, or percentage) for new column
• Time lag: Number of periods to compare against— how far back will we retrieve the data point? This value uses the time granularity of the underlying data
• Color bounds: Number bounds used for color encoding from red to blue. Reverse the numbers for blue to red. To get pure red or blue, you enter either only min or max
• Width: Width of sparkline graph
• Height: Height of sparkline graph
• Time ratio: Number of periods to ratio against
• Show Y-axis: Show Y-axis on the sparkline. Will display the manually set min/max if set or min/max values in the data otherwise
• Y-axis bounds: Manually set min/max for the Y-axis
• Number/Date format: Optional  d3 format string

Other Customizations

• URL: Templated link, it's possible to include {{ metric }} or other values coming from the controls