CODEWITHSUNDEEP
exceloptimizationvbabenchmarking
Gaijinhunter
Gaijinhunter

Reputation: 14685

Benchmarking VBA Code

What is considered the most accurate way to benchmark VBA code (in my case, I am testing code in Excel)? Are there any other techniques for benchmarking code besides the 2 below, and if so, what are the pros/cons of the method?

Here are 2 popular methods.

First: Timer

Sub TimerBenchmark()

Dim benchmark As Double
benchmark = Timer

'Do your code here

MsgBox Timer - benchmark

End Sub

And Tick (which I see argued as the most accurate):

Option Explicit
Private Declare Function GetTickCount Lib "kernel32" () As Long

Sub TickBenchmark()

Dim Start As Long
Dim Finish As Long

Start = GetTickCount()

'Do your code here

Finish = GetTickCount()
MsgBox CStr((Finish - Start) / 1000)

End Sub

Upvotes: 15

Views: 15527

Answers (5)

jonadv
jonadv

Reputation: 492

As RonnieDickson answered the QueryPerformanceCounter function is the most accurate possible way to bench VBA code (when you don't want to use a dll, like brettdj suggested). I wrote a class that makes that function easy to use: only initialize the benchmark class + call the method in between your code. No need to write code for substracting times, re-initializing times and writing to debug for example.

Sub TimerBenchmark()

Dim bm As New cBenchmark

'Some code here
bm.Wait 0.0001
bm.TrackByName "Some code"

'Some more (time consuming) code here
bm.Wait 0.04
bm.TrackByName "Bottleneck code"

End Sub

Which would automatically print a readable table to the Immediate window:

IDnr  Name             Count  Sum of tics  Percentage  Time sum
0     Fast code            1       25.136       5,91%   2,51 ms
1     Bottleneck code      1      400.316      94,09%     40 ms
      TOTAL                2      425.452     100,00%     43 ms

Total time recorded:             43 ms

The class includes a .Wait function, which does the same as Application.Wait, but requires only an input in seconds, instead of a time value. It is used above to simulate some code.

https://github.com/jonadv/VBA-Benchmark

Upvotes: 1

RonnieDickson
RonnieDickson

Reputation: 1420

The following code uses a windows function that is more accurate than Excel. It is taken from http://msdn.microsoft.com/en-us/library/aa730921.aspx#Office2007excelPerf_MakingWorkbooksCalculateFaster. The same page also contains some great tips on improving performance in Excel 2007.

Private Declare Function getFrequency Lib "kernel32" _
Alias "QueryPerformanceFrequency" (cyFrequency As Currency) As Long
Private Declare Function getTickCount Lib "kernel32" _
Alias "QueryPerformanceCounter" (cyTickCount As Currency) As Long

Function MicroTimer() As Double

  'Returns seconds.

  Dim cyTicks1 As Currency
  Static cyFrequency As Currency
  MicroTimer = 0

  ' Get frequency.
  If cyFrequency = 0 Then getFrequency cyFrequency

  ' Get ticks.
  getTickCount cyTicks1                            

  ' Seconds
  If cyFrequency Then MicroTimer = cyTicks1 / cyFrequency 
End Function

Upvotes: 18

brettdj
brettdj

Reputation: 55682

Professional Excel Development contains a dll utility PerfMon that I prefer, for its accuracy, and as it can be easily inserted throughout code with a couple of menu clicks

Upvotes: 1

Mike Dunlavey
Mike Dunlavey

Reputation: 40669

Any measurement is going to be noisy, so if you want precision, repeat the measurement many times and average the result.

Upvotes: 3

JMax
JMax

Reputation: 26591

Interesting question. This is not really a full answer but this is too long to be posted as a comment.
What i use is this kind of procedure:

Option Explicit
Public Time As Double
Sub Chrono()
If Time = 0 Then
    Time = Now()
Else
    MsgBox "Total time :" & Application.Text(Now() - _
        Time, "mm:ss") & "."  'or whatever if not a msgbox
    Time = 0
End If
End Sub

That way, you can put your code wherever you want and only have to call it twice (for instance):

If C_DEBUG Then Call Chrono

At the beginning and at the end of the part of code you want to test.

Yet, i would say there is no real "accurate" method because it also depends on what is running on your computer. I'd say these methods would mostly help telling which code is better than another.

Upvotes: 7

Related Questions