Syntax :

=FORECAST.ETS.CONFINT( target_date, values, timeline, [confidence_level], [seasonality], [data completion], [aggregation] )

Parameters :

• The function parameters are:
• target date –
• A date/time for which you wish to predict a value.
  (Must be after the final date/time in the historical timeline.)

Timeline –

• An independent array of dates/times corresponding to each value.
  This must meet the following requirements:
  The timeline array and the values array must have the same size.
  Although the dates/times in the timeline must have a constant step length between them:
  Depending on the value of the [data completeness] option, up to 30% of points may be missing and dealt with.
  The timeline may contain duplicates, whose associated values will be aggregated according to the [aggregation] option.
  The timeline’s dates and times can be in any sequence.

values –

• An array of previously known values for which you wish to anticipate the next point.

[confidence level] –

• An optional numeric number expressing the confidence level for the estimated confidence interval.
  If provided, this must be an integer between 0 and 1. (exclusive).
  If omitted, the [confidence level] is set to the default value of 0.95. (i.e. a 95 percent confidence level).

[seasonality] –

• An optional input specifying the algorithm to be used to detect seasonality in data.
  If provided, this must be a positive integer between 0 and 8784, corresponding to:

[data completion] –

• An optional input that specifies how the algorithm should handle missing timeline points. If provided, the [data completion] parameter can have a value of 0 or 1, which means:

[aggregation] –

• An optional parameter that tells the algorithm how to aggregate data with the same timestamp. If provided, this can be any integer between 1 and 7, which means: