Description
A statisticsSerial transformation will compute the configured statistic function from the list below. For each timestep in the output series the input values within the aggregation period are used to get a result value for that output time.
By default, the aggregation period is defined by the output timestep. A moving window can be defined by specifying an aggregationPeriod in the timeSeriesSet. In that case (see example below) for every timestep an aggregated value over de aggregationPeriod will be calculated.
Note: For the AggregationPeriod the start is exclusive and the end is inclusive.
Functions
The available statistic functions are:
- count
- countFlags
- kurtosis
- max
- mean (acts as movingaverage transformation)
- geometric mean (2019.01)
- median
- min
- percentileExceedence: (calculates the parameter value which is exceeded x% of the time in the aggregation period)
- percentileNonExceedence: (calculates the parameter value which is not exceeded x% of the time over the entire timeseries. The output of this percentileNonExceedence(x) function is what is colloquially known as the xth percentile of that timeseries.)
- quartile
- rootMeanSquareError
- rsquared
- skewness
- standardDeviation
- sum
- variance
Configuration Example
Input variable definition:
Transformation definition:
Output example
Description
A statisticsSerial transformation will compute the configured statistic function from the list below. For each timestep in the output series the input values within the aggregation period are used to get a result value for that output time.
By default, the aggregation period is defined by the output timestep. A moving window can be defined by specifying an aggregationPeriod in the timeSeriesSet. In that case (see example below) for every timestep an aggregated value over de aggregationPeriod will be calculated.
Note: For the AggregationPeriod the start is exclusive and the end is inclusive.
Functions
The available statistic functions are:
- count
- countFlags
- kurtosis
- max
- mean (acts as movingaverage transformation)
- geometric mean (2019.01)
- median
- min
- percentileExceedence
- percentileNonExceedence
- quartile
- rootMeanSquareError
- rsquared
- skewness
- standardDeviation
- sum
- variance
Trim Output
Note that this optional and default false, but very recommended to use to prevent overwriting existing values by missings (see logWarnExistingValueOverwrittenByMissing log messages). This problem will happen in case the transformation is running at a smaller cardinal timestep than the aggregation timestep. For example, running at 00:05, while aggregating to clock hours will result in a non-full clock hour at the start of the aggregation, so result in a missing value. With the trimOutput=true, that missing value will be ignored.
Configuration Example
Input variable definition:
Transformation definition:
Output example
