Atlanta Custom Software Development 

 
   Search        Code/Page
 

User Login
Email

Password

 

Forgot the Password?
Services
» Web Development
» Maintenance
» Data Integration/BI
» Information Management
Programming
  Database
Automation
OS/Networking
Graphics
Links
Tools
» Regular Expr Tester
» Free Tools

QuickSort - Exploiting the principle of exchanging keys

Total Hit ( 2131)

Rate this article:     Poor     Excellent 

 Submit Your Question/Comment about this article

Rating


 


Click here to copy the following block
' QuickSort. QuickSort, CombSort and ShellSort all exploit the principle of
' exchanging keys that are far apart in the list rather than adjacent. 
' QuickSort does this most elegantly and rapidly. The approach is to choose a
' "pivot" value (ideally, the median key) and then to work from each end of the
' list toward the middle. A key at each end is compared to the pivot and
' nothing is done if the left key is less than the pivot or the right key is
' greater. When a left key greater than the pivot and a right key less than
' the pivot have been found, those keys (or their pointers) are swapped,
' and the process continues until the left and right pointers cross. We then
' recursively call QuickSort on the left and right sublists until the lists are
' small (and delegate final sorting to low overhead InsertionSort).
'
' QuickSort does not need any auxiliary arrays, but uses a modest amount of
' stack space for recursion. It is not stable (although its descendent Ternary
' QuickSort is). On average, it is the fastest of the O(N log N) sorts,
' but it suffers from rare "worst case" behavior where certain input orders of
' keys cause speed to deteriorate to O(N^2). Naive implementations of
' QuickSort that choose the middle key for pivot exhibit O(N^2) behavior on
' sorted lists. The version of QuickSort presented here makes worst case
' behavior very unlikely by choosing the median of the first,
' last and middle keys as pivot. Two versions are provided. pQuickSortS is
' set up for strings and can be adapted to doubles by changing the declaration
' of array A(). QuickSortL is set up for longs, or A() can be redeclared for
' integers. 
'
' Reference: Robert Sedgewick, "Implementing Quicksort Programs",
' Comm. of the ACM 21(10):847-857 (1978).
'
' Speed: pQuickSortS sorts 500,000 random strings in 30.3 sec; sorts 100186
' library call numbers in 11.3 sec; sorts 25479 dictionary words in 2.0 sec
' (random order), 1.3 sec (presorted) or 1.8 sec (reverse sorted). QuickSortL
' sorts 500,000 random longs in 56 seconds. Timed in Excel 2001 on an 800 mhz
' PowerBook.
'
' Bottom line: contends with RadixSort for fastest; better adapted than Radix
' for non-string data, but not stable.

' Usage: 

Dim S1(L To R) As String
Dim P1(L To R) As Long
Dim L1(L To R) As Long

For I = L To R
  S1(I) = GetRandomString()
  P1(I) = I
  L1(I) = GetRandomLong()
Next I

pQuickSortS L, R, S1, P1
QuickSortL L, R, L1

' CODE:

Sub pQuickSortS(L As Long, R As Long, A() As String, P() As Long)
  'We put "sentinel" values flanking the real keys to avoid an extra test in
  ' the inner loop.
  A(L - 1) = MinStr
  A(R + 1) = MaxStr
  'We mostly sort the list with QuickSort.
  pQuickS L, R, A(), P
  'Then we finish up with low overhead InsertionSort
  pInsertS L, R, A(), P
End Sub

Sub pQuickS(L As Long, R As Long, A() As String, P() As Long)
  Dim MED As Long
  Dim LP As Long
  Dim RP As Long
  Dim Pivot As String
  Dim TMP As Long
  
  'Sublists <= 12 keys will be finished by running the whole list once thru
  ' InsertionSort.
  If R - L > 12 Then
  'Get the median pointer...
    MED = (L + R) \ 2
  'and swap it to the leftmost position.
    TMP = P(MED)
    P(MED) = P(L)
    P(L) = TMP
  'Now compare the leftmost, next leftmost & rightmost to choose a median of
  ' 3...
    If A(P(L + 1)) > A(P(R)) Then
      TMP = P(L + 1)
      P(L + 1) = P(R)
      P(R) = TMP
    End If
    If A(P(L)) > A(P(R)) Then
      TMP = P(L)
      P(L) = P(R)
      P(R) = TMP
    End If
    If A(P(L + 1)) > A(P(L)) Then
      TMP = P(L + 1)
      P(L + 1) = P(L)
      P(L) = TMP
    End If
  'and use its key as our pivot.
    Pivot = A(P(L))
  'Now work inward from each end.
    LP = L
    RP = R + 1
    Do
    'Scan right for a pointer whose key >= Pivot. In case Pivot is the
    ' largest key, we have
    'a sentinel value of MaxStr in A(R + 1) that will end a runaway loop. 
    ' Using the sentinel
    'avoids having a second test in the inner loop,
    ' so it can be as fast as possible.
      Do
        LP = LP + 1
      Loop While A(P(LP)) < Pivot
    'Scan left for a pointer whose key <= Pivot. Again,
    ' we have a sentinel value of MinStr
    'in A(L - 1) to stop the loop if Pivot is the smallest value in the
    ' list.
       Do
        RP = RP - 1
      Loop While A(P(RP)) > Pivot
    'If the pointers have crossed we're done.
      If RP < LP Then Exit Do
    'Otherwise, swap the pair we've identified.
      TMP = P(LP)
      P(LP) = P(RP)
      P(RP) = TMP
    Loop
  'Swap the pointer of the Pivot value back into place.
    TMP = P(L)
    P(L) = P(RP)
    P(RP) = TMP
  'Sort the shorter sublist first so the recursion stack is limited to
  ' logarithmic depth.
    If (RP - 1) - L < R - LP Then
      pQuickS L, RP - 1, A, P
      pQuickS LP, R, A, P
    Else
      pQuickS LP, R, A, P
      pQuickS L, RP - 1, A, P
    End If
  End If
End Sub

Sub pInsertS(L As Long, R As Long, A() As String, P() As Long)
  Dim LP As Long
  Dim RP As Long
  Dim TMP As Long
  Dim T As String
  
  For RP = L + 1 To R
    TMP = P(RP)
    T = A(TMP)
    For LP = RP To L + 1 Step -1
      If T < A(P(LP - 1)) Then P(LP) = P(LP - 1) Else Exit For
    Next LP
    P(LP) = TMP
  Next RP
End Sub

Sub QuickSortL(L As Long, R As Long, A() As Long)
  A(L - 1) = MinStr
  A(R + 1) = MaxStr
  QuickL L, R, A
  InsertL L, R, A
End Sub

Sub QuickL(L As Long, R As Long, A() As Long)
  Dim MED As Long
  Dim LP As Long
  Dim RP As Long
  Dim Pivot As String
  Dim TMP As Long
  
  If R - L > 12 Then
    MED = (L + R) \ 2
    TMP = A(MED)
    A(MED) = A(L)
    A(L) = TMP
    If A(L + 1) > A(R) Then
      TMP = A(L + 1)
      A(L + 1) = A(R)
      A(R) = TMP
    End If
    If A(L) > A(R) Then
      TMP = A(L)
      A(L) = A(R)
      A(R) = TMP
    End If
    If A(L + 1) > A(L) Then
      TMP = A(L + 1)
      A(L + 1) = A(L)
      A(L) = TMP
    End If
    Pivot = A(L)
    LP = L
    RP = R + 1
    Do
      Do
        LP = LP + 1
      Loop While A(LP) < Pivot
      Do
        RP = RP - 1
      Loop While A(RP) > Pivot
      If RP < LP Then Exit Do
      TMP = A(LP)
      A(LP) = A(RP)
      A(RP) = TMP
    Loop
    TMP = A(L)
    A(L) = A(RP)
    A(RP) = TMP
    If (RP - 1) - L < R - LP Then
      QuickL L, RP - 1, A
      QuickL LP, R, A
    Else
      QuickL LP, R, A
      QuickL L, RP - 1, A
    End If
  End If
End Sub

Sub InsertL(L As Long, R As Long, A() As Long)
  Dim LP As Long
  Dim RP As Long
  Dim TMP As Long
  
  For RP = L + 1 To R
    TMP = A(RP)
    For LP = RP To L + 1 Step -1
      If TMP < A(LP - 1) Then A(LP) = A(LP - 1) Else Exit For
    Next LP
    A(LP) = TMP
  Next RP
End Sub


Submitted By : Nayan Patel  (Member Since : 5/26/2004 12:23:06 PM)

Job Description : He is the moderator of this site and currently working as an independent consultant. He works with VB.net/ASP.net, SQL Server and other MS technologies. He is MCSD.net, MCDBA and MCSE. In his free time he likes to watch funny movies and doing oil painting.
View all (893) submissions by this author  (Birth Date : 7/14/1981 )


Home   |  Comment   |  Contact Us   |  Privacy Policy   |  Terms & Conditions   |  BlogsZappySys

© 2008 BinaryWorld LLC. All rights reserved.