rescale_fixed_width_position: Tripartite linear rescaling
In ndukler/tuSelecter2: Selecting transcript model using PRO-seq

View source: R/transcript_shape_profile-class.R

Given a length, a fixed bin width, and head/tail lengths this function generates a sequence [0, 1] that is composed of three linear sub-scales for the head, mid, and tail region. These scales are created such that positions <= linear_head_length are converted to a uniform sequence between [0, head_percent] and values [bin_size * length_out - linear_tail_length, bin_size] are converted to a uniform sequence [tail_percent, 1] with the mid values falling in the remaining range.

rescale_fixed_width_position(
  bin_size,
  length_out,
  linear_head_length,
  linear_tail_length,
  head_percent = 0.2,
  tail_percent = 0.2,
  short_heuristic = TRUE
)

`bin_size`	width of each position range that is being converted
`length_out`	length of position vector to be created
`linear_head_length`	total width of head region (# of bins * width)
`linear_tail_length`	total width of tail region (# of bins * width)
`head_percent`	fraction of [0, 1] to be occupied by head region
`tail_percent`	fraction of [0, 1] to be occupied by tail region
`short_heuristic`	whether sequences shorter than the sum of the head and tail length should be accommodated by proportionally rescaling the head and tail lengths. Otherwise an error will be thrown.