module Biocaml_histogram:Histograms with polymorphic bin types.sig
..end
A histogram is a list of bins, each with a count. Bin i
is defined by a lower limit lo(i)
and an upper limit hi(i)
. It is inclusive of the lower limit and exclusive of the upper limit. For all i
, hi(i) = lo(i+1)
. By convention the first bin is numbered 0. The count of a bin is a floating point number, allowing fractional values if necessary.
type 'a
t
'a
.val make : ('a -> 'a -> int) -> 'a list -> 'a t
make cmp bins
returns a new histogram from the given bins
, all initialized to a count of 0.0. The bins
must be provided as a list of the boundaries dividing them. The list [v0; v1; ...; vn]
of length n+1
represents the n
bins [v0, v1)
, [v1, v2)
, ..., [vn-1, vn)
, where cmp
is used as the comparison function. Raise Failure
if bins
are not monotonically increasing, or if length of bins
is less than 2.val to_list : 'a t -> (('a * 'a) * float) list
val copy : 'a t -> 'a t
val bin : 'a t -> int -> 'a * 'a
bin hist i
returns the i
th bin of hist
. Raise Failure
if an invalid bin number is requested.val count : 'a t -> int -> float
count hist i
returns the count the i
th bin. Raise Failure
if an invalid bin number is requested.val num_bins : 'a t -> int
val minimum : 'a t -> 'a
val maximum : 'a t -> 'a
val increment : ?delt:float -> 'a t -> 'a -> 'a t
increment delt hist x
increments the count of the bin containing x
by delt
(default is 1.0). The histogram is unaltered if x
not in any bin. This is not considered an error because it is often necessary to calculate a histogram for a subset of a larger data set.val reset : 'a t -> 'a t
val find_bin_index : 'a t -> 'a -> int option
find_bin_index hist x
returns the index of the bin in hist
containing x
. Return None if x
is outside the histogram's range.val in_range : 'a t -> 'a -> bool
in_range hist x
is true if x
greater than or equal to minimum hist
and strictly less than maximum hist
.val make_uniform : float -> float -> int -> float t
make_uniform min max n
returns a histogram with n
bins uniformly dividing up the range from min
to max
. Bins will be inclusive of the lower limit and exclusive of the upper limit, i.e. value of min
will fall into lowest bin and value of max
will fall outside the range of the histogram. Raise Failure
if min
not strictly less than max
or if n
not greater than 0.