rangenostics: Rangewise diagnostics
In sa-lee/superintronic: Tools for Working with Intron Signal

Rangewise diagnostics

lumpy(x, .size = .autosize(x))

bumpy(x, .size = .autosize(x))

crossing_points(x, .threshold = mean(x))

flat_spots(x, .threshold = 0, .tol = .Machine$double.eps^0.5)

count_grams(x, .size = .autosize(x))

max_mean_shift(x, .size, .step)

max_var_shift(x, .size, .step)

`x`	a numeric vector or Rle object
`.size`	the integer size of the window
`.threshold`	a numeric
`.tol`	is equal to .Machine$double^0.5
`.step`	the integer step size to shift the window

These are inspired by the concept of cognostics by John Tukey, and related work in the area by Lee Wilkinson (graph theoretic scagnostics) and in time series analysis by Rob Hyndman.

The

sa-lee/superintronic documentation built on Feb. 18, 2020, 10:36 a.m.