This function returns the frequency of a time series. In the case of a sparse xts() time series, and in other cases, the R functions xts::periodicity() and frequency() do not return BIMETS compliant values. Therefore, these functions have been extended.
## S3 method for class 'xts'frequency(x,...)
Arguments
x: Input time series.
...: Backward compatibility.
Returns
This function returns the integer value stored in the attribute .bimetsFreq of the input time series, if any. Otherwise, the frequency will be calculated by using the shortest time difference between two observations, while accounting for day-saving and bissextile years.
See Also
normalizeYP
NUMPERIOD
BIMETS indexing
Examples
#build a sparse xts() xArr<-rnorm(13) dateArr<-seq(as.Date('2000/01/01'),by='6 months',length=10) dateArr2<-seq(as.Date('2010/01/01'),by='3 months',length=3)#strange array of dates here below... dateArr3<-c(dateArr,dateArr2) dataF<-data.frame(dateArr3,xArr) xt<-xts(dataF[,2],order.by=dataF[,1])#get bimets calculated frequency cat(frequency(xt))#print 4... without bimets R returns 1#...legacy periodicity() periodicity(xt)