Visual Basic for Applications/A PRNG for VBA
Summary[edit | edit source]
- A pseudo-random number generator (PRNG), if run for long enough, generates a characteristic sequence that is based on its algorithm. This sequence repeats forever and is invariant. The Rnd() function of VBA, if placed in a loop without a parameter, and without making use of Randomize() at all, will generate 16,777,216 values between zero and unity, then start again at the beginning, making a repeating sequence with a cycle length of 16,777,216. The only option for the user is to choose the starting point within that one sequence. This is done by choosing a start value or seed. In the case of Rnd(), the start value is chosen in two ways: by default, using the system timer, or with a user-set number. Again, the start parameters that are set by the user do not make new sequences, they just decide which bit of that one long sequence will be used. Linear Congruential Generators (LCG), the type used by Microsoft's Rnd() function are described in detail at Linear congruential generator.
- The maximum cycle length for a single stage LCG is equal to its modulus. For combined generators, the maximum cycle length is equal to the least common multiple of the cycle lengths of the individual generators. A well-designed generator will have the maximum cycle length and consist only of unique values throughout its sequence, but not all generators are well-designed. The above link describes the design values required to make an LCG with a maximum cycle length over all of its starting values.
- The code module below contains the Wichmann-Hill (1982) CLCG (combined LCG) in VBA and is fully functional. It is called RndX() and is used in conjunction with its own RandomizeX(). It has a much longer repeat cycle than Microsoft's Rnd(). A summary of the most useful settings for RndX() is given, with additional details for those who need them in a drop box. Sadly, this author lacks the tools and knowledge for any serious testing of number generators, so the offerings below are likely to be of interest only to beginners.
- Long-cycle generators are awkward to study in Excel. The cycles of both Microsoft's Rnd() and the user function RndX() are much too long to write a complete cycle to a single worksheet column. The solution is either to list only parts of long streams or to make a number generator with a cycle short enough for a full listing. Listing in a single column this way allows confirmation of the repeat cycle length, then after trimming the rows to a complete set, counting rows after the removal of duplicates will confirm for the skeptical that all of the values are unique. A module is included with procedures to list one section of the Wichmann-Hill implementation, that fits into about 30269 rows or so, and another with a very simple generator for further testing, that fits into just 43.
Microsoft's Rnd() algorithm[edit | edit source]
Microsoft's Visual Basic for Applications (VBA), at present uses a linear congruential generator (LCG) for pseudo-random number generation in the Rnd() function. Attempts to implement the Microsoft algorithm in VBA failed owing to overflow. The following is its basic algorithm.
x1 = ( x0 * a + c ) MOD m and; Rnd() = x1/m where: Rnd() = returned value m = modulus = (2^24) x1 = new value x0 = previous value (initial value 327680) a = 1140671485 c = 12820163 Repeat length = m = (2^24) = 16,777,216
Similarities will be noticed between Microsoft's Rnd() and the one below, described by Wichmann-Hill (1982), in which a sum of three LCG expressions is used in the production of each output number. The combination of expressions gives RndX(), with the coded values, its much improved cycle length of:
Cycle length = least_common_multiple(30268, 30306, 30322) = 30268 * 30306 * 30322 / 4 = 6,953,607,871,644
VBA Code - Wichmann-Hill (1982)[edit | edit source]
A reminder about module level variables may be in order. Module level variables hold their values between procedure runs. In fact they will retain values until the VBA is no longer used at all or the code is edited. The code has been laced with resets for these variables, to ensure starting with intended values, as opposed to old stored ones from the previous top procedure runs.
On a cautionary note; although this algorithm has improved properties over the resident Rnd(), the applications on which these generators are run are not particularly secure. Consider also that the output of all LCG coding is entirely predictable if the starting value is ever known. In fact, if any part of such a stream is known, then it is possible for those who intend harm to find the entire stream by comparing it with stored values. These facts when taken together limit the use of such a VBA implementation to study or non-critical applications.
That said, these are likely to be the most useful parameter configurations: In each case RandomizeX() should only be called once, before and outside any generator loop that contains RndX(). This advice also applies to the Microsoft function Rnd() and its companion Randomize().
- To produce outputs with an unpredictable start point, and a different start point each time it is run:
- Call RandomizeX without any parameter before calling RndX, also without any parameter. This uses the system timer.
- To produce outputs from a large set of start points, repeatable, and chosen by a user parameter:
- Call RandomizeX with any numeric parameter before calling RndX without any parameter. Changed RandomizeX parameter values result in different start points of the standard algorithm stream.
- To produce an unpredictable, single value, different each time it is run:
- Call RandomizeX without any parameter before calling RndX with a parameter of zero. This uses the system timer.
- To produce a repeatable single value, related to, and chosen by a user parameter:
- Call RandomizeX with any numeric parameter before calling RndX with a parameter of zero. Changed RandomizeX parameter values result in different values that are peculiar to each parameter.
- Refer to the drop box below for a complete tabulation of the parameter settings and their outcomes.
The code in this section should be saved as a separate standard module in Excel.
Option Explicit Dim nSamples As Long Dim nX As Long, nY As Long, nZ As Long Sub TestRndX() 'run this to obtain RndX() samples 'Wichmann, Brian; Hill, David (1982), Algorithm AS183: 'An Efficient and Portable Pseudo-Random Number Generator, 'Journal of the Royal Statistical Society. Series C Dim n As Long 'reset module variables nX = 0: nY = 0: nZ = 0 RandomizeX For n = 1 To 10 Debug.Print RndX() MsgBox RndX() Next n 'reset module variables nX = 0: nY = 0: nZ = 0 End Sub Sub TestScatterChartOfPRNG() 'run this to make a point scatter chart 'using samples from RndX Dim vA As Variant, n As Long Dim nS As Long, nR As Double 'remove any other charts 'DeleteAllCharts 'reset module variables nX = 0: nY = 0: nZ = 0 'set number of samples here nSamples = 1000 ReDim vA(1 To 2, 1 To nSamples) 'dimension array 'load array with PRNG samples RandomizeX For n = 1 To nSamples nR = RndX() vA(1, n) = n 'x axis data - sample numbers vA(2, n) = nR 'y axis data - prng values Next n 'make scatter point chart from array ChartScatterPoints vA, 1, 2, nSamples & " Samples of RndX()", _ "Sample Numbers", "PRNG Values [0,1]" 'reset module work variables nX = 0: nY = 0: nZ = 0 End Sub Sub RandomizeX(Optional ByVal nSeed As Variant) 'sets variables for PRNG procedure RndX() Const MaxLong As Double = 2 ^ 31 - 1 Dim nS As Long Dim nN As Double 'make multiplier If IsMissing(nSeed) Then nS = Timer * 60 Else nN = Abs(Int(Val(nSeed))) If nN > MaxLong Then 'no overflow nN = nN - Int(nN / MaxLong) * MaxLong End If nS = nN End If 'update variables nX = (nS Mod 30269) nY = (nS Mod 30307) nZ = (nS Mod 30323) 'avoid zero state If nX = 0 Then nX = 171 If nY = 0 Then nY = 172 If nZ = 0 Then nZ = 170 End Sub Function RndX(Optional ByVal nSeed As Long = 1) As Double 'PRNG - gets pseudo random number - use with RandomizeX 'Wichmann-Hill algorithm of 1982 Dim nResult As Double 'initialize variables If nX = 0 Then nX = 171 nY = 172 nZ = 170 End If 'first update variables If nSeed <> 0 Then If nSeed < 0 Then RandomizeX (nSeed) nX = (171 * nX) Mod 30269 nY = (172 * nY) Mod 30307 nZ = (170 * nZ) Mod 30323 End If 'use variables to calculate output nResult = nX / 30269# + nY / 30307# + nZ / 30323# RndX = nResult - Int(nResult) End Function Sub ChartScatterPoints(ByVal vA As Variant, RowX As Long, RowY As Long, _ Optional sTitle As String = "", Optional sXAxis As String, _ Optional sYAxis As String) 'array input must contain two data rows for x and y data 'parameters for user title, x axis and y axis labels 'makes a simple point scatter chart Dim LBC As Long, UBC As Long, LBR As Long, UBR As Long, n As Long, bOptLim As Boolean Dim X As Variant, Y As Variant, sX As String, sY As String, sT As String, oC As Chart LBR = LBound(vA, 1): UBR = UBound(vA, 1) LBC = LBound(vA, 2): UBC = UBound(vA, 2) ReDim X(LBC To UBC) ReDim Y(LBC To UBC) 'labels for specific charts If sTitle = "" Then sT = "Title Goes Here" Else sT = sTitle If sXAxis = "" Then sX = "X Axis Label Goes Here" Else sX = sXAxis If sYAxis = "" Then sY = "Y Axis Label Goes Here" Else sY = sYAxis If RowX < LBR Or RowX > UBR Or RowY < LBC Or RowY > UBC Then MsgBox "Parameter data rows out of range in ChartColumns - closing" Exit Sub End If 'transfer data to chart arrays For n = LBC To UBC X(n) = vA(RowX, n) 'x axis data Y(n) = vA(RowY, n) 'y axis data Next n 'make chart Charts.Add 'set chart type ActiveChart.ChartType = xlXYScatter 'point scatter chart 'remove unwanted series With ActiveChart Do Until .SeriesCollection.Count = 0 .SeriesCollection(1).Delete Loop End With 'assign the data and labels to a series With ActiveChart.SeriesCollection If .Count = 0 Then .NewSeries If Val(Application.Version) >= 12 Then .Item(1).Values = Y .Item(1).XValues = X Else .Item(1).Select Names.Add "_", X ExecuteExcel4Macro "series.x(!_)" Names.Add "_", Y ExecuteExcel4Macro "series.y(,!_)" Names("_").Delete End If End With 'apply title string, x and y axis strings, and delete legend With ActiveChart .HasTitle = True .ChartTitle.Text = sT .SetElement (msoElementPrimaryCategoryAxisTitleAdjacentToAxis) 'X .Axes(xlCategory).AxisTitle.Text = sX .SetElement (msoElementPrimaryValueAxisTitleRotated) 'Y .Axes(xlValue).AxisTitle.Text = sY .Legend.Delete End With 'trim axes to suit With ActiveChart 'X Axis .Axes(xlCategory).Select .Axes(xlCategory).MinimumScale = 0 .Axes(xlCategory).MaximumScale = nSamples .Axes(xlCategory).MajorUnit = 500 .Axes(xlCategory).MinorUnit = 100 Selection.TickLabelPosition = xlLow 'Y Axis .Axes(xlValue).Select .Axes(xlValue).MinimumScale = -0.2 .Axes(xlValue).MaximumScale = 1.2 .Axes(xlValue).MajorUnit = 0.1 .Axes(xlValue).MinorUnit = 0.05 End With ActiveChart.ChartArea.Select Set oC = Nothing End Sub Sub DeleteAllCharts5() 'run this to delete all ThisWorkbook charts Dim oC Application.DisplayAlerts = False For Each oC In ThisWorkbook.Charts oC.Delete Next oC Application.DisplayAlerts = True End Sub
Simpler Tests of PRNGs[edit | edit source]
The code module below contains a stripped down version of the Wichmann-Hill (1982) algorithm, in fact using only the first of its three calculated sections. It will make several complete streams of values on Sheet1 of the workbook in which it is run, using different start values. Notice that the first values are all repeated at row 30269, as will the whole stream if extended. After producing the list, use the spreadsheet's functions for column sorting and the removal of duplicates to see that each column contains the appropriate number of unique entries. An even simpler generator with a repeat cycle of just 43 is also included that might make study more manageable, and the cycle of Microsoft's Rnd() can be seen to repeat at 16777216 (+1) by running TestMSRnd.
The code in this section should be saved as a separate standard module in Excel.
Option Explicit Private ix2 As Long Sub TestWHRnd30269() 'makes five columns for complete output streams 'each with a different start point 'runs a simplified LCNG with mod 30269 Dim sht As Worksheet, nS As Double, nSamp As Long Dim c As Long, r As Long, nSeed As Long 'set seed value for Rnd2() nSeed = 327680 'WH initial seed 'set number of random samples to make nSamp = 30275 '30269 plus say, 6 'set initial value of carry variable ix2 = nSeed Set sht = ThisWorkbook.Worksheets("Sheet1") 'clear the worksheet sht.Cells.Cells.ClearContents 'load sheet with set of samples For c = 1 To 5 'number of runs ix2 = nSeed + c 'change start value For r = 1 To nSamp 'number of samples nS = WHRnd30269() 'get a sample sht.Cells(r, c) = nS 'write to sheet Next r Next c sht.Cells(1, 1).Select End Sub Function WHRnd30269() As Double 'first part of Wichmann-Hill tripple. 'When started with seed ix2 = 171, 'full sequence repeats from n = 30269 'without any repeated values before. Dim r As Double 'ix2 cannot be 0. If ix2 = 0 Then ix2 = 171 End If 'calculate Xn+1 from Xn ix2 = (171 * ix2) Mod 30269 'make an output value r = ix2 / 30269# WHRnd30269 = r - Int(r) End Function Sub TestSimpleRnd43() 'makes five columns for complete output streams 'each with a different start point 'runs a very simple LCNG with mod 43 Dim sht As Worksheet, nS As Double, nSamp As Long Dim c As Long, r As Long, nSeed As Long 'set seed value for Rnd2() nSeed = 17 'initial seed 'set number of random samples to make nSamp = 45 '43 plus say, 2 'set initial value of carry variable ix2 = nSeed Set sht = ThisWorkbook.Worksheets("Sheet1") 'clear the worksheet sht.Cells.Cells.ClearContents 'load sheet with set of samples For c = 1 To 5 'number of runs ix2 = nSeed + c 'change start value For r = 1 To nSamp 'number of samples nS = SimpleRnd43() 'get a sample sht.Cells(r, c) = nS 'write to sheet Next r Next c sht.Cells(1, 1).Select End Sub Function SimpleRnd43() As Double 'simple Lehmer style LCNG to show repeat streams 'produces one sequence of 42 unique values - then repeats entire sequence 'start value decides only where the predictable sequence begins Dim r As Double 'Note; Makes 42 unique values before sequence repeats 'Modulus = 43: Multiplier = 5: Initial Seed = 17 '43 is prime '5 is primitive root mod 43 '17 is coprime to 43 'ix2 cannot be 0. If ix2 = 0 Then ix2 = 17 End If 'calculate a new carry variable ix2 = (5 * ix2) Mod 43 'make an output value r = ix2 / 43# SimpleRnd43 = r - Int(r) End Function Sub TestMSRnd() 'makes two sets of single data using MS Rnd 'the first 10 samples of Rnd() values 'followed by values around sample 16777216 'confirms sequence probably re-starts at M+1 = 16777217 Dim sht As Worksheet, nS As Double Dim c As Long, r As Long, nMod As Long 'note modulus nMod = 16777216 Set sht = ThisWorkbook.Worksheets("Sheet1") 'clear the worksheet sht.Cells.Cells.ClearContents 'load sheet with set of samples For r = 1 To nMod + 20 'number of samples nS = Rnd() 'get a sample Select Case r Case 1 To 10 sht.Cells(r, 1) = r sht.Cells(r, 2) = nS Case (nMod - 4) To (nMod + 5) sht.Cells(r - 16777211 + 10, 1) = r sht.Cells(r - 16777211 + 10, 2) = nS End Select Next r sht.Cells(1, 1).Select End Sub
References[edit | edit source]
- Wichmann, Brian; Hill, David (1982), Algorithm AS183: An Efficient and Portable Pseudo-Random Number Generator, Journal of the Royal Statistical Society. Series C
See Also[edit | edit source]
- Wichmann-Hill CLCG: A Wikipedia article on the specific combined linear congruential generator in question.
- Linear congruential generator: A good Wikipedia description of the conditions for a full cycle.
External Links[edit | edit source]
- How Visual Basic Generates Pseudo-Random Numbers for the RND Function: Microsoft kb231847 knowledge base item