Run the plotting-position and threshold cascade over an analysis_data() frame and return the
record diagnostics. The plotting positions are the Hirsch-Stedinger censored positions, so
interval and threshold observations shift the positions of the systematic record even though
they carry no plotting ordinate themselves.
Value
A named list. index, value, plotting_position, and is_low_outlier are parallel
vectors over the exact series; number_of_low_outliers, low_outlier_threshold,
plotting_parameter, lambda (events per index), total_record_length,
zero_value_relative_frequency, and the four series counts are scalars.
Examples
peaks <- c(12500, 15300, 8900, 22100, 18700, 14200, 9800, 28500)
s <- analysis_data_summary(analysis_data(peaks))
data.frame(value = s$value, position = s$plotting_position)
#> value position
#> 1 12500 0.6666667
#> 2 15300 0.4444444
#> 3 8900 0.8888889
#> 4 22100 0.2222222
#> 5 18700 0.3333333
#> 6 14200 0.5555556
#> 7 9800 0.7777778
#> 8 28500 0.1111111
# A perception threshold reweights the systematic record's positions.
d <- analysis_data(
exact = peaks,
threshold = data.frame(start_index = 8, end_index = 47, value = 25000, number_above = 0)
)
analysis_data_summary(d)$plotting_position
#> [1] 0.63281250 0.38802083 0.87760417 0.14322917 0.26562500 0.51041667 0.75520833
#> [8] 0.01041667