what i needed is…(1) in the first iteration i want 10. (2) in the
second iteration i want 5, 15. (3) in the third iteration i want
3,8,13,18. and so on. – user4365176
I found a tool (in my libraries) which I used to transfer the contents of a vector (of non-decreasing integers) TO a binary tree SUCH-THAT the simple binary tree would be (mostly) balanced.
Here is a version of that code which I simplified for your purpose. This snippet is a method in a class, so the undefined variables prefixed with ‘m_’ (m_tree, m_ss, m_R) are data attributes of that class. m_tree is instance of my home-grown-simple binary tree, with method “insertVal(size_t i)”, m_ss is a pointer to a std::stringstream, and m_R is the user input number, the Range.
‘genSeqRR()’ is recursive. It chops a range in half, then processes the right then the left (somewhat in the form of depth-first-in-order. Currently, the sequence is perhaps slightly biased. Not sure how to describe. See Note 2 (in code).
I also inserted these sequences to an AVL tree (from rosettacode.org/wiki) and the result were disappointing. Possibly not wrong, but not useful here.
You might try this insert sequence with a RedBlack Trees.
example outputs after the code:
// Generate a Sequence of size_t values over a fixed limited Range,
// Recursively using binary-pattern through the range, to achieve approximate
// balance when inserting into a simple binary tree.
//
// Note1: 'index' is range (0..N-1), value to insert is (indx+1) i.e 1..N
// Note2: prefer to insert bigger indx first
//
void genSeqRR(size_t depth, const size_t si, const size_t bi)
{
size_t rng = bi - si; // big indx - small indx
switch (rng)
{
default: // 3 or more elements, insert mi, and recurse
{
size_t delta = rng / 2;
size_t mi = si + delta; // mid index
(void)m_tree->insertVal(mi+1); // Note1
// if(m_ss)
*m_ss << " rng:" << std::setw(3) << rng
<< " dpth:" << std::setw(3) << depth // depth of genSeqRR
<< " iVal:" << std::setw(3) << mi+1
<< std::endl;
// Note2
genSeqRR (depth+1, mi+1, bi); // recurse on max - (middle + 1) thru biggest index
genSeqRR (depth+1, si, mi-1); // recurse on min - smallest index thru (middle - 1)
}
break;
case 0: // 1 elment
{
(void)m_tree->insertVal (si+1); // Note1
// if(m_ss)
*m_ss << " rng:__1"
<< " dpth:" << std::setw(3) << depth
<< " iVal:" << std::setw(3) << si+1 << std::endl;
}
break;
case 1: // 2 consecutive elements
{
// Note1, Note2
(void)m_tree->insertVal(bi+1);
(void)m_tree->insertVal(si+1);
// if(m_ss)
*m_ss << " rng:__2"
<< " dpth:" << std::setw(3) << depth
<< " iVal:" << std::setw(3) << si+1 << std::setw(3) << bi+1 << std::endl;
}
break;
}// switch
} // void genSeqRR()
with range m_R = 20,
genSeqRR (1, 0, (m_R-1)); // start recursion over whole range
output:
Anchor->showTallView()
'1'
'2'
'3'
'4'
'5'
'6'
'7'
'8'
'9'
'10'
'11'
'12'
'13'
'14'
'15'
'16'
'17'
'18'
'19'
'20'
with range m_R = 19,
genSeqRR (1, 0, (m_R-1)); // start recursion over whole range
output:
Anchor->showTallView()
'1'
'2'
'3'
'4'
'5'
'6'
'7'
'8'
'9'
'10'
'11'
'12'
'13'
'14'
'15'
'16'
'17'
'18'
'19'
with range m_R = 19,
Anchor->showTallView()
'1'
'2'
'3'
'4'
'5'
'6'
'7'
'8'
'9'
'10'
'11'
'12'
'13'
'14'
'15'
'16'
'17'
'18'