Cyclic Analysis (Time Series)
cyclicAnalysis()
Derive a set of cyclic statistics from the current regular interval time series data
set. The time series data set must have a time interval of “1HOUR”, “1DAY” or
“1MONTH.” The function sorts the time series values into statistical "bins"
relevant to the time interval. Values for the 1HOUR interval data are sorted into
24 bins representing the hours of the day, 0100 to 2400. The 1DAY interval data is
apportioned to 365 bins for the days of the year. The 1MONTH interval data is
sorted into 12 bins for the months of the year.
The format of the resultant data sets is as a “pseudo” time series for the year 3000. For example, the cyclic analysis of one month of hourly interval data will produce pseudo time series data sets having 24 hourly values for the day January 1, 3000. If the statistical parameter is the “maximum” value, then the 24 values represent the maximum value occurring at that hour of the day in the current time series. The cyclic analysis of daily interval data will produce pseudo time series data sets having 365 daily values for the year 3000. The cyclic analysis of monthly interval data will result in pseudo time series data sets having 12 monthly values for the year 3000.
Fourteen pseudo time series data sets are derived by the cyclic analysis function for the following statistical parameters:
§ Number of values processed for each time interval
§ Maximum value
§ Time of maximum value
§ Minimum value
§ Time of minimum value
§ Average value
§ Probability exceedence percentiles for 5%, 10%, 25%, 50% (median value), 75%, 90%, and 95%
§ Standard deviation
The 14 pseudo time series of cyclic statistics are returned by the function as an array of time series data sets. The parameter part of the record path for each time series is modified to indicate the type of the statistical parameter. For a flow record, the parameter "FLOW" would become "FLOW-MAX" for the maximum values statistics, "FLOW-P5" for the 5% percentile statistics, etc.
Parameters: Takes no parameters
Example: cyclicData = tsData.cyclicAnalysis()
Returns: A sequence of 14 TimeSeriesMath objects, each of which is a pseudo time series data sets representing a statistical parameter.
Generated Exceptions: Throws an hec.hecmath.HecMathException if the time series is not regular interval or does not have a time interval of "1HOUR”, "1DAY", or "1MONTH”.