LOCS function

Select Time Series Indices

This function returns the indices of the input TRUE elements. The input can be either an array or a time series. The result is usually used as a structured index to produce a new array.

LOCS(x=NULL, options='ALL', ...)

Arguments

• x: This function accepts as input a boolean array or a boolean time series, often as the result of a logic comparison between an expression and a numerical array or a numerical time series:

  e.g. LOCS(c(a,b,c,...)>=k); LOCS(ts==j);

  LOCS(ts<expr); LOCS(is.na(ts)); etc...

• options: A selection option can refine the result:

  ALL : (default) all the TRUE indices will be returned in the output.

  UNIQUE : return the index of the unique TRUE result; if there are multiple TRUE results then an error will be thrown.

  FIRST : return the first TRUE result.

  LAST : return the last TRUE result.

• ...: Backward compatibility.

Returns

This function returns a numerical array built with the indices of the values that are TRUE in the input boolean array or in the input boolean time series.

See Also

NOELS

is.bimets

BIMETS indexing

TSERIES

GETYEARPERIOD

NOELS

NAMELIST

INTS

TSINFO

TABIT

ELIMELS

Examples

#create ts
        n<-10
        ts1<-TSERIES((1:n),START=c(2000,1),FREQ=1)
        print(LOCS(ts1>7,options='FIRST')) #print 8
        
        #generate error: print LOCS(): input has more than one TRUE element.
        tryCatch({print(LOCS(ts1>=3,options='UNIQUE'));},error=function(e){print(e$message);})
        
        print(LOCS(is.na(c(1,2,NA,4,5,6,7,NA,NA)))) #print c(3,8,9)