unsort

emlib.combinatorics.unsort(seq, entropy, margin=0, tolerance=0.05, numiter=100, timeout=None)[source]

Generate a permutation of xs unsorted according to the given entropy.

Parameters:

  • seq (Sequence) – a sequence to be unsorted

  • entropy (float) – 0=the original sequence is returned; 1=random sequence is returned

  • margin – a number or tuple (left, right). These elements are left untouched

  • numiter – the number of times the algorithm is run. The best result will be returned

  • timeout (Optional[float]) – alternatively, you can specify a timeout (numiter will be disregarded)

Return type:

ndarray

Returns:

un unsorted version of seq, as numpy array, or None if

  • If entropy == 0: the original sequence is returned

  • If entropy == 1: a sequence is generated which is as random as possible (this does not mean that there cannot be any fixed points, it refers to the general result)

Examples

# unsort the first 10 numbers, leave 0 and 9 untouched at their places
unsort(range(10), 0.5, margin=1)

# unsort the given seq., do not touch the first too elements
unsort((1,3, 5, 4, 0), 0.2, margin=(2, 0))