series_record {RecordTest}

From Record Times to Time Series

Description

If only the record times are available (upper or lower, or both) and not the complete series, series_record builds a complete series with the same record occurrence as specified in the arguments. This function is useful to apply the plots and tests within RecordTest-package to a vector of record times.

Usage

series_record(L_upper, R_upper, L_lower, R_lower, Trows = NA)

Arguments

L_upper, L_lower

A vector of (increasing) integers denoting the upper or/and lower record times.

R_upper, R_lower

(Optional) A vector of (increasing/decreasing) values denoting the upper or/and lower record values.

Trows

Integer indicating the actual length of the series. If it is not specified, then the length of the series is assumed equal to the last record occurrence.

Value

A vector of length Trows with L_upper upper or/and L_lower lower record times and R_upper upper or/and R_lower lower record values.

Note

Remember that the first observation in a series is always a record time.

Author(s)

Jorge Castillo-Mateo

See Also

series_double, series_rev, series_split, series_ties, series_uncor, series_untie

Examples

# upper record times observed in a 100 length time series
L <- c(1, 4, 14, 40, 45, 90)
X <- series_record(L_upper = L, Trows = 100)

# now you can apply plots and tests for upper records to the X series
#N.plot(X)
#N_normal.test(X)

# if you also have lower record times
L_lower <- c(1, 2, 12, 56, 57, 78, 91)
X <- series_record(L_upper = L, L_lower = L_lower, Trows = 100)

# now you can apply plots and tests to the X series with both types of record times
#foster.plot(X, statistic = 'd')
#foster.test(X, statistic = 'd')

# apply to the 200-meter Olympic records from 1900 to 2020
or200m <- series_record(L_lower = Olympic_records_200m$t, 
                        R_lower = Olympic_records_200m$value,
                        Trows = 27)
# some plots and tests                    
N.plot(or200m, record = c(0,1,0,0))                         
N.test(or200m, record = "lower", distribution = "poisson-binomial")