Add a MultiIndexSet method to make an instance downward-closed

As part of the refactoring and update of the MultiIndexSet class (see Issue #99 (closed)), especially in relation to the separation between the notion of completeness and downward-closedness (see Issue #105 (closed)), a method to make an instance of MultiIndexSet downward-closed should be added to the class. This method (possibly called make_downward_closed() is parallel to the method make_complete() (see Issue #103 (closed)).

Here's an example of how it might be used:

>>> import numpy as np
>>> import minterpy as mp
>>> exponents = np.array([
... [0, 1, 0],
... [1, 0, 1],
... ])
>>> mi = mp.MultiIndexSet(exponents, lp_degree=1.0)
>>> mi.is_downward_closed  # See Issue #105
False
>>> mi_downward_closed = mi.make_downward_closed()  # By default, outplace modification
>>> mi_downward_closed
MultiIndexSet
[[0 0 0]  # new element
 [1 0 0]  # new element
 [0 1 0]
 [0 0 1]  # new element
 [1 0 1]]
>>> mi_downward_closed.is_downward_closed
True

Here, three new elements are added to the multi-index set such that the resulting set downward-closed. The resulting set is sorted lexicographically.

Similar to make_complete(), the current instance may be modified in-place as follows:

>>> mi.make_downward_closed(inplace=True)
>>> mi.is_downward_closed
True
>>> mi
MultiIndexSet
[[0 0 0]  # new element
 [1 0 0]  # new element
 [0 1 0]
 [0 0 1]  # new element
 [1 0 1]]

As discussed in Issue #105 (closed), downward-closed multi-index set may not be complete. For instance, the complete multi-index set of three-dimensional polynomial with respect to l_p-degree 1.0 having a polynomial degree of:

>>> mi.poly_degree
2

is:

>>> mp.MultiIndexSet.from_degree(3, 2, 1.0)
MultiIndexSet
[[0 0 0]
 [1 0 0]
 [2 0 0]
 [0 1 0]
 [1 1 0]
 [0 2 0]
 [0 0 1]
 [1 0 1]
 [0 1 1]
 [0 0 2]]

From the above complete set there are many several possible downward-closed sets; The DDS algorithm only requires a downward-closed set.