Trait indxvec::Vecops

source · [−]
pub trait Vecops<T> {

Show 35 methods    fn maxt(self) -> T
    where
        T: PartialOrd + Copy;
    fn mint(self) -> T
    where
        T: PartialOrd + Copy;
    fn minmaxt(self) -> (T, T)
    where
        T: PartialOrd + Copy;
    fn minmax(self) -> MinMax<T>
    where
        T: PartialOrd + Copy;
    fn minmax_slice(self, i: usize, n: usize) -> MinMax<T>
    where
        T: PartialOrd + Copy;
    fn minmax_indexed(self, idx: &[usize], i: usize, n: usize) -> MinMax<T>
    where
        T: PartialOrd + Copy;
    fn revs(self) -> Vec<T>
    where
        T: Copy;
    fn sansrepeat(self) -> Vec<T>
    where
        T: PartialEq + Copy;
    fn member(self, m: T) -> Option<usize>
    where
        T: PartialEq + Copy;
    fn memsearch(self, val: T) -> Option<usize>
    where
        T: PartialOrd;
    fn memsearchdesc(self, val: T) -> Option<usize>
    where
        T: PartialOrd;
    fn memsearch_indexed(self, i: &[usize], val: T) -> Option<usize>
    where
        T: PartialOrd;
    fn memsearchdesc_indexed(self, i: &[usize], val: T) -> Option<usize>
    where
        T: PartialOrd;
    fn binsearch(self, val: T) -> usize
    where
        T: PartialOrd;
    fn binsearchdesc(self, val: T) -> usize
    where
        T: PartialOrd;
    fn occurs(self, val: T) -> usize
    where
        T: PartialOrd;
    fn occurs_multiple(self, sdesc: &[T], val: T) -> usize
    where
        T: PartialOrd + Copy;
    fn unite_unsorted(self, v: &[T]) -> Vec<T>
    where
        T: Clone;
    fn intersect(self, v2: &[T]) -> Vec<T>
    where
        T: PartialOrd + Copy;
    fn intersect_indexed(self, ix1: &[usize], v2: &[T], ix2: &[usize]) -> Vec<T>
    where
        T: PartialOrd + Copy;
    fn diff(self, v2: &[T]) -> Vec<T>
    where
        T: PartialOrd + Copy;
    fn diff_indexed(self, ix1: &[usize], v2: &[T], ix2: &[usize]) -> Vec<T>
    where
        T: PartialOrd + Copy;
    fn partition(self, pivot: T) -> (Vec<T>, Vec<T>, Vec<T>)
    where
        T: PartialOrd + Copy;
    fn partition_indexed(self, pivot: T) -> (Vec<usize>, Vec<usize>, Vec<usize>)
    where
        T: PartialOrd + Copy;
    fn merge(self, v2: &[T]) -> Vec<T>
    where
        T: PartialOrd + Copy;
    fn merge_indexed(
        self, 
        idx1: &[usize], 
        v2: &[T], 
        idx2: &[usize]
    ) -> (Vec<T>, Vec<usize>)
    where
        T: PartialOrd + Copy;
    fn merge_indices(self, idx1: &[usize], idx2: &[usize]) -> Vec<usize>
    where
        T: PartialOrd + Copy;
    fn mergesort(self, i: usize, n: usize) -> Vec<usize>
    where
        T: PartialOrd + Copy;
    fn sortidx(self) -> Vec<usize>
    where
        T: PartialOrd + Copy;
    fn sortm(self, ascending: bool) -> Vec<T>
    where
        T: PartialOrd + Copy;
    fn rank(self, ascending: bool) -> Vec<usize>
    where
        T: PartialOrd + Copy;
    fn isorttwo(self, idx: &mut [usize], i0: usize, i1: usize) -> bool
    where
        T: PartialOrd;
    fn isortthree(self, idx: &mut [usize], i0: usize, i1: usize, i2: usize)
    where
        T: PartialOrd;
    fn hashsort_indexed(self, min: f64, max: f64) -> Vec<usize>
    where
        T: PartialOrd + Copy,
        f64: From<T>;
    fn hashsortslice(
        self, 
        idx: &mut [usize], 
        i: usize, 
        n: usize, 
        min: f64, 
        max: f64
    )
    where
        T: PartialOrd + Copy,
        f64: From<T>;

}
Expand description

Methods to manipulate Vecs

Required Methods

Maximum value in self

Minimum value in self

Minimum and maximum values in self

Returns MinMax{min, minindex, max, maxindex}

MinMax of n items starting at subscript i

MinMax of a subset of self, defined by its idx subslice between i,i+n.

Reversed copy of self

Repeated items removed

Some(subscript) of the first occurence of m, or None

Binary search for the subscript of the first occurence of val

Binary search for the subscript of the last occurence of val

Binary search for val via ascending sort index i

Backwards binary search for val via descending sort index i

Binary search of an explicitly sorted list in ascending order. Returns an index of the first item that is greater than val. When none are greater, returns s.len()

Binary search of an explicitly sorted list in descending order. Returns an index of the first item that is smaller than val. When none are smaller, returns s.len()

Counts occurrences of val by simple linear search of an unordered set

Efficiently counts number of occurences from ascending and descending sorts

Unites (concatenates) two unsorted sets. For union of sorted sets, use merge

Intersects two ascending explicitly sorted generic vectors.

Intersects two ascending index sorted vectors.

Removes items of sorted v2 from sorted self.

Removes items of v2 from self using their sort indices.

Divides an unordered set into three: items smaller than pivot, equal, and greater

Divides an unordered set into three by the pivot. The results are subscripts to self

Merges (unites) two sorted sets, result is also sorted

Merges (unites) two sets, using their sort indices, giving also the resulting sort index

Used by merge_indexed

Utility used by sortidx

Stable Merge sort main method, giving sort index

Stable Merge sort, explicitly sorted result obtained via sortidx

Rank index obtained via sortidx

Utility, swaps any two items into ascending order

Utility, sorts any three items into ascending order

Stable Hash sort

Utility used by hashsort_indexed

Implementations on Foreign Types

Maximum value T of slice &[T]

Minimum value T of slice &[T]

Minimum and maximum (T,T) of a slice &[T]

Minimum, minimum’s first index, maximum, maximum’s first index

Finds min and max of a subset of self, defined by its subslice between i,i+n. Returns min of self, its index, max of self, its index.

Using only a subset of self, defined by its idx subslice between i,i+n. Returns min of self, its index’s index, max of self, its index’s index.

Reverse a generic slice by reverse iteration. Creates a new Vec. Its naive use for descending sort etc. is to be avoided for efficiency reasons.

Removes repetitions from an explicitly ordered set.

Finds the first occurence of item m in self by iteration. Returns Some(index) to the slice or None (when it has gone to the end). Note that it uses only partial order and thus accepts any item that is neither greater nor smaller than m (equality by default). Suitable for small unordered sets. For longer lists or repeated membership tests, it is better to index sort them and then use faster binary memsearch (see below).

Binary search of an explicitly sorted list (in ascending order). Returns Some(index) of any item that is equal to val. When none are found, returns None. Example use: membership of an ascending ordered set.

Binary search of an explicitly sorted list (in descending order). Returns Some(index) of any item that is neither smaller nor greater than val. When none are found, returns None. Example use: membership of an descending ordered set.

Binary search of an indexed list (in ascending order). Returns Some(index) of any item that is neither smaller nor greater than val. When none are found, returns None. Example use: membership of an indexed ordered set.

Binary search of an indexed list (in descending order). Returns Some(index) of any item that is neither smaller nor greater than val. When none are found, returns None. Example use: membership of an indexed descending set.

Binary search of an explicitly sorted list in ascending order. Returns an index of the first item that is greater than val. When none are greater, returns s.len() (invalid index but logical). The complement index (the result subtracted from s.len()), gives the first item in descending order that is not greater than val. Note that both complements of binsearch and binsearchdesc, in their respective opposite orderings, refer to the same preceding item iff there exists precisely one item equal to val. However, there can be more than one such items or none. Example use: looking up cummulative probability density functions.

Binary search of an explicitly sorted list in descending order. Returns an index of the first item that is smaller than val. When none are smaller, returns s.len() (invalid index but logical). The complement index (the result subtracted from s.len()), gives the first item in ascending order that is not smaller than val. Note that both complements of binsearch and binsearchdesc, in their respective opposite orderings, refer to the same preceding item iff there exists precisely one item equal to val. However, there can be more than one such items or none. Example use: looking up cummulative probability density functions.

Counts occurrences of val by simple linear search of any unordered set

Counts occurrences of val, using previously obtained ascending explicit sort sasc and descending sort sdesc. The two sorts must be of the same original set! This is to facilitate counting of many different values without having to repeat the sorting. This function is efficient at counting numerous repetitions in large sets (e.g. probabilities in stats). Binary search from both ends is deployed: O(2log(n)).

Example:
use crate::indxvec::Indices;
use indxvec::Vecops;
let s = [1.,2.,3.14159,3.14159,4.,5.,6.];
let sindx = s.sortidx(); // only one sorting
let sasc = sindx.unindex(&s,true);   // explicit ascending
let sdesc = sindx.unindex(&s,false); // explicit descending
assert_eq!(sasc.occurs_multiple(&sdesc,3.14159),2);

Unites (joins) two unsorted sets. For union of sorted sets, use merge

Intersects two ascending explicitly sorted generic vectors.

Intersects two ascending index-sorted generic vectors. Returns a single explicitly ordered set.

Sets difference: deleting elements of the second from the first. Two ascending explicitly sorted generic vectors.

Sets difference: deleting elements of the second from the first. Two ascending index sorted generic vectors.

Partition with respect to a pivot into three sets

Partition by pivot gives three sets of indices.

Merges two explicitly ascending sorted generic vectors, by classical selection and copying of their head items into the result. Consider using merge_indexed instead, especially for non-primitive end types T.

Merges two ascending sort indices. Data is not shuffled at all, v2 is just concatenated onto v1 in one go and both remain in their original order. Returns the concatenated vector and a new valid sort index into it.

Merges the sort indices of two concatenated vectors. Data in s is not changed at all, only consulted for the comparisons. This function is used by mergesort and merge_indexed.

Doubly recursive non-destructive merge sort. The data is not moved or mutated. Efficiency is comparable to quicksort but more stable Returns a vector of indices to s from i to i+n, such that the indexed values are in ascending sort order (a sort index). Only the index values are being moved.

A wrapper for mergesort, to obtain the sort index of the (whole) input vector. Simpler than sortm.

Immutable sort. Returns new sorted vector (ascending or descending). Is a wrapper for mergesort. Passes the boolean flag ‘ascending’ onto ‘unindex’. Mergesort by itself always produces only an ascending index.

Fast ranking of many T items, with only n*(log(n)+1) complexity. Ranking is done by inverting the sort index. Sort index is in sorted order, giving data positions. Ranking is in data order, giving sorted order positions. Thus sort index and ranks are in an inverse relationship. They are easily converted by .invindex() (for: invert index).

swap any two index items, if their data items (self) are not in ascending order

sort three index items if their self items are out of ascending order

N recursive non-destructive hash sort. Input data are read only. Output is sort index. Requires min,max, the data range, that must enclose all its values. The range is often known. If not, it can be obtained with minmaxt().

Implementors