Centered Moving Average
Smoothing
centeredMovingAverage(integer
numberToAverageOver,
boolean
onlyValidValues,
boolean
useReduced)
Derive a new time series from the centered moving average of numberToAverageOver
values in the current time series. numberToAverageOver must be an odd integer
greater than 2.
If onlyValidValues is set to true, then if any points in the averaging interval
are missing, the point in the new time series is set to missing. If
onlyValidValues is set to false and missing values are contained in the averaging
interval, a smoothed point is still computed using the remaining valid values in the
interval. If there are no valid values in the averaging interval, the point is set
to missing.
If useReduced is set to true, then centered moving average points can still be
computed at the beginning and end of the time series, even if there are less than
numberToAverageOver values in the averaging interval. If useReduced
is set to false, then the first and last numberToAverageOver/2 points of the
resultant time series are set to missing.
Parameters:
numberToAverageOver – An integer containing the number of values to average
over for computing the centered moving average. Must be odd and greater than 2.
onlyValidValues – Either Constants.TRUE, or Constants.FALSE, specifying
whether all values in the averaging interval must be valid for the computed point in the
new time series to be valid.
useReduced – Either Constants.TRUE, or Constants.FALSE, specifying
whether to allow points at the beginning and end of the resultant time series to be
computed from a reduced ( less than numberToAverageOver ) set of points.
Example:
avgData = tsData.centeredMovingAverage (5,
Constants.TRUE,
Constants.TRUE)
Returns: A new TimeSeriesMath object
Generated Exceptions: Throws
a HecMathException if the numberToAverageOver is less than 3 or not odd.