A chart may show trends based on data from other chart's dataset. There are four types of trends:

  • Linear Regression
  • Exponential Regression
  • Moving Average
  • Percentile

Trending involves two charts to present: the source chart (its data is used to calculate the trend) and the trend chart itself. Thus, trends can be plotted by the following charts only when they are a part of Mixed XY Chart:

To configure new trend for a chart, perform the following:

  • Create a Mixed XY Chart
  • Add a "simple" XY Chart to this mixed chart.
  • Configure the data source for the simple XY chart and ensure that it shows correct data.
  • Add another XY chart to the mixed chart. The second XY chart will show trend(s) of first XY chart's source data.
  • Change the Data Source Type of second XY chart to Trending.
  • Since trend chart and source chart share the same address space and axes, you can now remove trend chart's "original" axes from the parent mixed chart. This step is optional.
  • Open Trending properties of the second XY chart and change Source Chart Name to the name of widget component matching the "source" XY chart, e.g. xyLineChart1.
  • Select trend type and configure other trend properties.
  • Run the widget and test the trends.

    • The trending properties are available only for XY Line, Area and Bar charts when they are added as sub-charts to a Mixed XY Chart.

    Source Chart

    Name of sub-chart component those data will be used to build trend curves. It must be the a sub-chart of the same Mixed XY Chart with this trending sub-chart.

    Property name: sourceDataPlotID

    Property type: String

    Type of trend(s) to build: Linear, Exponential, Moving Average, or Percentile.

    Property name: trendingType

    Property type: Integer

    Forecast Factor

    Length of "future" trend in time comparing to the time frame of the source data, for example:

    • Value of 0.0 means that the trend is build exactly for the time frame covered by source data.
    • Value of 0.5 means that the trend goes into the future for half of source data's time frame. E.g. if the source data covers the whole year 2012, the trend will cover 2012 and first two quarters of 2013.
    • Value of 1.0 means that the trend goes into the future full source data's time frame. E.g. if the source data covers the whole year 2012, the trend will cover 2012 and 2013.
    • Value of 3.0 means that the trend goes into the future three source time frames, etc.

    Property name: forecastFactor

    Property type: Float

    Linear and Exponential Trend Options

    Point Count

    Number of data items in the trend curve.

    Zero value means that point count will match the number of data items in the source series.

    This option is valid for Linear and Exponential trends only.

    Property name: pointCount

    Property type: Float

    Moving Average Trend Options

    Period

    The averaging period for Moving Average calculation (in chart's data space).

    Property name: period

    Property type: Float

    Skip Period

    The length of the initial skip period for the Moving Average calculation (in chart's data space).

    Property name: skipPeriod

    Property type: Float

    Percentile Trend Options

    Percentile

    The base for Percentile trend, in percents. The default is value is 95%, making percentile trend appear so that 95% of source values will stay below the trend line.

    Property name: percentile

    Property type: Float

    Was this page helpful?