Time a function
You are encouraged to solve this task according to the task description, using any language you may know.
- Task
Write a program which uses a timer (with the least granularity available on your system) to time how long a function takes to execute.
Whenever possible, use methods which measure only the processing time used by the current process; instead of the difference in system time between start and finish, which could include time used by other processes on the computer.
This task is intended as a subtask for Measure relative performance of sorting algorithms implementations.
8051 Assembly
Using a timer requires knowledge on two things: the oscillator frequency (which limits the maximum precision) and the desired precision. This code uses a common crystal of 11.0592MHz - but provides values for a few others as examples. This code also uses a precision of 4 bits (2^(-4) = 0.0625 seconds). Those familiar with binary can think of this as a right shift of 4 of the multi-byte value, where the low 4 bits represent the fraction of a second, and the remaining bits represent whole seconds. The maximum time value depends on the number of bytes used and the precision. For x bytes and p precision, the maximum value you can count to is (256^x - 1) * 2^(-p).
<lang asm>TC EQU 8 ; number of counter registers TSTART EQU 08h ; first register of timer counter TEND EQU TSTART + TC - 1 ; end register of timer counter
- Note
- The multi-byte value is stored in Big-endian
- Some timer reloads
_6H EQU 085h ; 6MHz _6L EQU 0edh _12H EQU 00bh ; 12MHz _12L EQU 0dbh _110592H EQU 01eh ; 11.0592MHz _110592L EQU 0ffh
- How to calculate timer reload (e.g. for 11.0592MHz)
- Note
- 1 machine cycle takes 12 oscillator periods
- 11.0592MHz / 12 * 0.0625 seconds = 57,600 cycles = e100h
- ffffh - e100h = NOT e100h = 1effh
- assuming a 11.0592MHz crystal
TIMERH EQU _110592H TIMERL EQU _110592L
- some timer macros (using timer0)
start_timer macro setb tr0 endm stop_timer macro clr tr0 endm reset_timer macro mov tl0, #TIMERL mov th0, #TIMERH endm
increment_counter macro ;; increment counter (multi-byte increment) push psw push acc push 0 ; r0 mov r0, #TEND+1 setb c inc_reg: dec r0 clr a addc a, @r0 mov @r0, a jnc inc_reg_ ; end prematurally if the higher bytes are unchanged cjne r0, #TSTART, inc_reg inc_reg_: ; if the carry is set here then the multi byte value has overflowed pop 0 pop acc pop psw endm
ORG RESET jmp init ORG TIMER0 jmp timer_0
timer_0: ; interrupt every 6.25ms stop_timer ; we only want to time the function reset_timer increment_counter start_timer reti
init: mov sp, #TEND setb ea ; enable interrupts setb et0 ; enable timer0 interrupt mov tmod, #01h ; timer0 16-bit mode reset_timer
; reset timer counter registers clr a mov r0, #TSTART clear: mov @r0, a inc r0 cjne r0, #TEND, clear
start_timer call function ; the function to time stop_timer
; at this point the registers from TSTART ; through TEND indicate the current time ; multiplying the 8/16/24/etc length value by 0.0625 (2^-4) gives ; the elapsed number of seconds ; e.g. if the three registers were 02a0f2h then the elapsed time is: ; 02a0f2h = 172,274 and 172,274 * 0.0625 = 10,767.125 seconds ; ; Or alternatively: ; (high byte) 02h = 2 and 2 * 2^(16-4) = 8192 ; (mid byte) a0h = 160 and 160 * 2^(8-4) = 2560 ; (low byte) f2h = 242 and 242 * 2^(0-4) = 15.125 ; 8192 + 2560 + 15.125 = 10,767.125 seconds
jmp $
function: ; do whatever here ret
END </lang>
ACL2
<lang Lisp>(time$ (nthcdr 9999999 (take 10000000 nil)))</lang>
Output (for Clozure):
; (EV-REC *RETURN-LAST-ARG3* ...) took ; 2.53 seconds realtime, 2.48 seconds runtime ; (160,001,648 bytes allocated). (NIL)
Ada
<lang ada>with Ada.Calendar; use Ada.Calendar; with Ada.Text_Io; use Ada.Text_Io;
procedure Query_Performance is
type Proc_Access is access procedure(X : in out Integer); function Time_It(Action : Proc_Access; Arg : Integer) return Duration is Start_Time : Time := Clock; Finis_Time : Time; Func_Arg : Integer := Arg; begin Action(Func_Arg); Finis_Time := Clock; return Finis_Time - Start_Time; end Time_It; procedure Identity(X : in out Integer) is begin X := X; end Identity; procedure Sum (Num : in out Integer) is begin for I in 1..1000 loop Num := Num + I; end loop; end Sum; Id_Access : Proc_Access := Identity'access; Sum_Access : Proc_Access := Sum'access;
begin
Put_Line("Identity(4) takes" & Duration'Image(Time_It(Id_Access, 4)) & " seconds."); Put_Line("Sum(4) takes:" & Duration'Image(Time_It(Sum_Access, 4)) & " seconds.");
end Query_Performance;</lang>
Example
Identity(4) takes 0.000001117 seconds. Sum(4) takes: 0.000003632 seconds.
Aime
<lang aime>integer identity(integer x) {
return x;
}
integer
sum(integer c)
{
integer s;
s = 0; while (c) {
s += c; c -= 1;
}
return s;
}
real
time_f(integer (*fp) (integer), integer fa)
{
date f, s; time t;
d_now(s);
fp(fa);
d_now(f);
t_ddiff(t, f, s);
return t_microsecond(t) / 1000000r;
}
integer
main(void)
{
o_real(6, time_f(identity, 1)); o_text(" seconds\n"); o_real(6, time_f(sum, 1000000)); o_text(" seconds\n");
return 0;
}</lang>
AutoHotkey
System time
Uses system time, not process time <lang AutoHotkey>MsgBox % time("fx") Return
fx() {
Sleep, 1000
}
time(function, parameter=0) {
SetBatchLines -1 ; don't sleep for other green threads StartTime := A_TickCount %function%(parameter) Return ElapsedTime := A_TickCount - StartTime . " milliseconds"
}</lang>
Using QueryPerformanceCounter
QueryPerformanceCounter allows even more precision: <lang AHK>MsgBox % time("fx")
time(function, parameter=0){ SetBatchLines -1 DllCall("QueryPerformanceCounter", "Int64*", CounterBefore) DllCall("QueryPerformanceFrequency", "Int64*", Freq) %function%(parameter) DllCall("QueryPerformanceCounter", "Int64*", CounterAfter) return (CounterAfter-CounterBefore)/Freq * 1000 " milliseconds" }
fx(){ Sleep 1000 }</lang>
BaCon
The BaCon TIMER function keeps track of time spent running, in milliseconds (which is also the time unit used by SLEEP). This is not process specific, but a wall clock time counter which starts at 0 during process initialization. As BaCon can easily use external C libraries, process specific CLOCK_PROCESS_CPUTIME_ID clock_gettime could also be used.
<lang freebasic>' Time a function SUB timed()
SLEEP 7000
END SUB
st = TIMER timed() et = TIMER PRINT st, ", ", et</lang>
- Output:
prompt$ ./time-function 0, 7000
BASIC
<lang qbasic>DIM timestart AS SINGLE, timedone AS SINGLE, timeelapsed AS SINGLE
timestart = TIMER SLEEP 1 'code or function to execute goes here timedone = TIMER
'midnight check: IF timedone < timestart THEN timedone = timedone + 86400 timeelapsed = timedone - timestart</lang>
See also: BBC BASIC, PureBasic.
Batch File
Granularity: hundredths of second. <lang Batch File> @echo off Setlocal EnableDelayedExpansion
call :clock
- timed function:fibonacci series.....................................
set /a a=0 ,b=1,c=1
- loop
if %c% lss 2000000000 echo %c% & set /a c=a+b,a=b, b=c & goto loop
- ....................................................................
call :clock
echo Function executed in %timed% hundredths of second goto:eof
- clock
if not defined timed set timed=0 for /F "tokens=1-4 delims=:.," %%a in ("%time%") do ( set /A timed = "(((1%%a - 100) * 60 + (1%%b - 100)) * 60 + (1%%c - 100)) * 100 + (1%%d - 100)- %timed%" ) goto:eof </lang>
BBC BASIC
<lang bbcbasic>start%=TIME:REM centi-second timer REM perform processing lapsed%=TIME-start%</lang>
Bracmat
<lang bracmat>( ( time
= fun funarg t0 ret . !arg:(?fun.?funarg) & clk$:?t0 & !fun$!funarg:?ret & (!ret.flt$(clk$+-1*!t0,3) s) )
& ( fib
= . !arg:<2&1 | fib$(!arg+-1)+fib$(!arg+-2) )
& time$(fib.30) )</lang> Output:
1346269.5,141*10E0 s
C
On some system (like GNU/Linux) to be able to use the clock_gettime function you must link with the rt (RealTime) library.
CLOCK_PROCESS_CPUTIME_ID
is preferred when available (eg. Linux kernel 2.6.12 up), being CPU time used by the current process. (CLOCK_MONOTONIC
generally includes CPU time of unrelated processes, and may be drifted by adjtime()
.)
<lang c>#include <stdio.h>
- include <time.h>
int identity(int x) { return x; }
int sum(int s) {
int i; for(i=0; i < 1000000; i++) s += i; return s;
}
- ifdef CLOCK_PROCESS_CPUTIME_ID
/* cpu time in the current process */
- define CLOCKTYPE CLOCK_PROCESS_CPUTIME_ID
- else
/* this one should be appropriate to avoid errors on multiprocessors systems */
- define CLOCKTYPE CLOCK_MONOTONIC
- endif
double time_it(int (*action)(int), int arg) {
struct timespec tsi, tsf;
clock_gettime(CLOCKTYPE, &tsi); action(arg); clock_gettime(CLOCKTYPE, &tsf);
double elaps_s = difftime(tsf.tv_sec, tsi.tv_sec); long elaps_ns = tsf.tv_nsec - tsi.tv_nsec; return elaps_s + ((double)elaps_ns) / 1.0e9;
}
int main() {
printf("identity (4) takes %lf s\n", time_it(identity, 4)); printf("sum (4) takes %lf s\n", time_it(sum, 4)); return 0;
}</lang>
C++
<lang cpp>#include <ctime>
- include <iostream>
using namespace std;
int identity(int x) { return x; } int sum(int num) {
for (int i = 0; i < 1000000; i++) num += i; return num;
}
double time_it(int (*action)(int), int arg) {
clock_t start_time = clock(); action(arg); clock_t finis_time = clock(); return ((double) (finis_time - start_time)) / CLOCKS_PER_SEC;
}
int main() {
cout << "Identity(4) takes " << time_it(identity, 4) << " seconds." << endl; cout << "Sum(4) takes " << time_it(sum, 4) << " seconds." << endl; return 0;
}</lang>
Example
Identity(4) takes 0 seconds. Sum(4) takes 0.01 seconds.
C#
Using Stopwatch.
<lang csharp>using System; using System.Linq; using System.Threading; using System.Diagnostics;
class Program {
static void Main(string[] args) { Stopwatch sw = new Stopwatch();
sw.Start(); DoSomething(); sw.Stop();
Console.WriteLine("DoSomething() took {0}ms.", sw.Elapsed.TotalMilliseconds); }
static void DoSomething() { Thread.Sleep(1000);
Enumerable.Range(1, 10000).Where(x => x % 2 == 0).Sum(); // Sum even numers from 1 to 10000 }
}</lang>
Using DateTime.
<lang csharp>using System; using System.Linq; using System.Threading;
class Program {
static void Main(string[] args) { DateTime start, end;
start = DateTime.Now; DoSomething(); end = DateTime.Now;
Console.WriteLine("DoSomething() took " + (end - start).TotalMilliseconds + "ms"); }
static void DoSomething() { Thread.Sleep(1000);
Enumerable.Range(1, 10000).Where(x => x % 2 == 0).Sum(); // Sum even numers from 1 to 10000 }
}</lang>
Output:
DoSomething() took 1071,5408ms
Clojure
<lang clojure>
(defn fib [] (map first (iterate (fn a b [b (+ a b)]) [0 1])))
(time (take 100 (fib)))
</lang>
Output:
"Elapsed time: 0.028 msecs" (0 1 1 2 3 5 8 13 21 34 55 89 144 233 377 610 987 1597 2584 4181)
Common Lisp
Common Lisp provides a standard utility for performance measurement, time:
<lang lisp>CL-USER> (time (reduce #'+ (make-list 100000 :initial-element 1))) Evaluation took:
0.151 seconds of real time 0.019035 seconds of user run time 0.01807 seconds of system run time 0 calls to %EVAL 0 page faults and 2,400,256 bytes consed.</lang>
(The example output here is from SBCL.)
However, it merely prints textual information to trace output, so the information is not readily available for further processing (except by parsing it in a CL-implementation-specific manner).
The functions get-internal-run-time and get-internal-real-time may be used to get time information programmatically, with at least one-second granularity (and usually more). Here is a function which uses them to measure the time taken for one execution of a provided function:
<lang lisp>(defun timings (function)
(let ((real-base (get-internal-real-time)) (run-base (get-internal-run-time))) (funcall function) (values (/ (- (get-internal-real-time) real-base) internal-time-units-per-second) (/ (- (get-internal-run-time) run-base) internal-time-units-per-second))))
CL-USER> (timings (lambda () (reduce #'+ (make-list 100000 :initial-element 1)))) 17/500 7/250</lang>
D
<lang d>import std.stdio, std.datetime;
int identity(int x) {
return x;
}
int sum(int num) {
foreach (i; 0 .. 100_000_000) num += i; return num;
}
double timeIt(int function(int) func, int arg) {
StopWatch sw; sw.start(); func(arg); sw.stop(); return sw.peek().usecs / 1_000_000.0;
}
void main() {
writefln("identity(4) takes %f6 seconds.", timeIt(&identity, 4)); writefln("sum(4) takes %f seconds.", timeIt(&sum, 4));
}</lang>
Output:
identity(4) takes 0.0000016 seconds. sum(4) takes 0.522065 seconds.
Using Tango
<lang d> import tango.io.Stdout; import tango.time.Clock;
int identity (int x) {
return x;
}
int sum (int num) {
for (int i = 0; i < 1000000; i++) num += i; return num;
}
double timeIt(int function(int) func, int arg) {
long before = Clock.now.ticks; func(arg); return (Clock.now.ticks - before) / cast(double)TimeSpan.TicksPerSecond;
}
void main () {
Stdout.format("Identity(4) takes {:f6} seconds",timeIt(&identity,4)).newline; Stdout.format("Sum(4) takes {:f6} seconds",timeIt(&sum,4)).newline;
} </lang>
E
— E has no standardized facility for CPU time measurement; this
.
<lang e>def countTo(x) { println("Counting...") for _ in 1..x {} println("Done!") }
def MX := <unsafe:java.lang.management.makeManagementFactory> def threadMX := MX.getThreadMXBean() require(threadMX.isCurrentThreadCpuTimeSupported()) threadMX.setThreadCpuTimeEnabled(true)
for count in [10000, 100000] { def start := threadMX.getCurrentThreadCpuTime() countTo(count) def finish := threadMX.getCurrentThreadCpuTime() println(`Counting to $count takes ${(finish-start)//1000000}ms`) }</lang>
Elena
ELENA 4.x : <lang elena>import system'calendar; import system'routines; import system'threading; import system'math; import extensions;
someProcess() {
threadControl.sleep(1000); new Range(0,10000).filterBy:(x => x.mod:2 == 0).summarize();
}
public program() {
var start := now; someProcess(); var end := now; console.printLine("Time elapsed in msec:",(end - start).Milliseconds)
}</lang>
- Output:
Time elapsed in msec:1015
Elixir
tc/1 <lang elixir>iex(10)> :timer.tc(fn -> Enum.each(1..100000, fn x -> x*x end) end) {236000, :ok}</lang> tc/2 <lang elixir>iex(11)> :timer.tc(fn x -> Enum.each(1..x, fn y -> y*y end) end, [1000000]) {2300000, :ok}</lang> tc/3 <lang elixir>iex(12)> :timer.tc(Enum, :to_list, [1..1000000]) {224000,
[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, ...]}</lang>
Erlang
Erlang's timer module has three implementations of the tc function.
tc/1 takes a 0-arity function and executes it: <lang erlang> 5> {Time,Result} = timer:tc(fun () -> lists:foreach(fun(X) -> X*X end, lists:seq(1,100000)) end). {226391,ok} 6> Time/1000000. % Time is in microseconds. 0.226391 7> % Time is in microseconds. </lang> tc/2 takes an n-arity function and its arguments: <lang erlang> 9> timer:tc(fun (X) -> lists:foreach(fun(Y) -> Y*Y end, lists:seq(1,X)) end, [1000000]). {2293844,ok} </lang> tc/3 takes a module name, function name and the list of arguments to the function: <lang erlang> 8> timer:tc(lists,seq,[1,1000000]). {62370,
[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22, 23,24,25,26,27|...]}
</lang>
Euphoria
<lang euphoria>atom t t = time() some_procedure() t = time() - t printf(1,"Elapsed %f seconds.\n",t)</lang>
F#
The .Net framework provides a Stopwatch class which provides a performance counter. <lang fsharp> open System.Diagnostics let myfunc data =
let timer = new Stopwatch() timer.Start() let result = data |> expensive_processing timer.Stop() printf "elapsed %d ms" timer.ElapsedMilliseconds result
</lang>
Factor
<lang factor>[ 10000 iota sum drop ] time</lang> Output:
Running time: 0.002888635 seconds Additional information was collected. dispatch-stats. - Print method dispatch statistics gc-events. - Print all garbage collection events gc-stats. - Print breakdown of different garbage collection events gc-summary. - Print aggregate garbage collection statistics
Forth
<lang forth>: time: ( "word" -- )
utime 2>R ' EXECUTE utime 2R> D- <# # # # # # # [CHAR] . HOLD #S #> TYPE ." seconds" ;
1000 time: MS \ 1.000081 seconds ok</lang>
Fortran
version 4.4.5 (Debian 4.4.5-8) on x86_64-linux-gnu
<lang fortran> c The subroutine to analyze
subroutine do_something()
c For testing we just do nothing for 3 seconds
call sleep(3) return end
c Main Program
program timing integer(kind=8) start,finish,rate call system_clock(count_rate=rate) call system_clock(start)
c Here comes the function we want to time
call do_something() call system_clock(finish) write(6,*) 'Elapsed Time in seconds:',float(finish-start)/rate return end
</lang>
FreeBASIC
<lang freebasic>' FB 1.05.0 Win64
Function sumToLimit(limit As UInteger) As UInteger
Dim sum As UInteger = 0 For i As UInteger = 1 To limit sum += i Next Return sum
End Function
Dim As Double start = timer Dim limit As UInteger = 100000000 Dim result As UInteger = sumToLimit(limit) Dim ms As UInteger = Int(1000 * (timer - start) + 0.5) Print "sumToLimit("; Str(limit); ") = "; result Print "took "; ms; " milliseconds to calculate" Print Print "Press any key to quit" Sleep</lang>
- Output:
sumToLimit(100000000) = 5000000050000000 took 314 milliseconds to calculate
GAP
<lang gap># Return the time passed in last function time;</lang>
Go
go test
The Go command line tool go test
includes benchmarking support.
Given a package with functions:
<lang go>package empty
func Empty() {}
func Count() {
// count to a million for i := 0; i < 1e6; i++ { }
}</lang> the following code, placed in a file whose name ends in _test.go, will time them: <lang go>package empty
import "testing"
func BenchmarkEmpty(b *testing.B) {
for i := 0; i < b.N; i++ { Empty() }
}
func BenchmarkCount(b *testing.B) {
for i := 0; i < b.N; i++ { Count() }
}</lang>
go test
varies b.N
to get meaningful resolution.
Example:
$ go test -bench=. testing: warning: no tests to run PASS BenchmarkEmpty 2000000000 0.30 ns/op BenchmarkCount 10000 298734 ns/op ok 3.642s
The first number is the value of b.N
chosen and the second the average time per iteration.
The testing
package can optionally include memory use and throughput benchmarks.
There is also a standard tool to compare the multiple benchmark outputs (installable via go get golang.org/x/tools/cmd/benchcmp
).
testing.Benchmark
The benchmarking features of the testing
package are exported for use within a Go program.
<lang go>package main
import (
"fmt" "testing"
)
func empty() {}
func count() {
for i := 0; i < 1e6; i++ { }
}
func main() {
e := testing.Benchmark(func(b *testing.B) { for i := 0; i < b.N; i++ { empty() } }) c := testing.Benchmark(func(b *testing.B) { for i := 0; i < b.N; i++ { count() } }) fmt.Println("Empty function: ", e) fmt.Println("Count to a million:", c) fmt.Println() fmt.Printf("Empty: %12.4f\n", float64(e.T.Nanoseconds())/float64(e.N)) fmt.Printf("Count: %12.4f\n", float64(c.T.Nanoseconds())/float64(c.N))
}</lang>
- Output:
Empty function: 2000000000 0.80 ns/op Count to a million: 2000 796071 ns/op Empty: 0.7974 Count: 796071.6555
Alternative technique
The go test
command and the testing
package are the preferred techniques for benchmarking or timing any Go code. Ignoring the well-tested and carefully crafted standard tools though, here is a simplistic alternative:
As the first line of the function you wish to time, use defer with an argument of time.Now() to print the elapsed time to the return of any function. For example, define the function from as shown below. It works because defer evaluates its function's arguments at the time the function is deferred, so the current time gets captured at the point of the defer. When the function containing the defer returns, the deferred from function runs, computes the elapsed time as a time.Duration, and prints it with standard formatting, which adds a nicely scaled unit suffix. <lang go>package main
import (
"fmt" "time"
)
func from(t0 time.Time) {
fmt.Println(time.Now().Sub(t0))
}
func empty() {
defer from(time.Now())
}
func count() {
defer from(time.Now()) for i := 0; i < 1e6; i++ { }
}
func main() {
empty() count()
}</lang> Output:
2us 643us
Groovy
CPU Timing
<lang groovy>import java.lang.management.ManagementFactory import java.lang.management.ThreadMXBean
def threadMX = ManagementFactory.threadMXBean assert threadMX.currentThreadCpuTimeSupported threadMX.threadCpuTimeEnabled = true
def clockCpuTime = { Closure c ->
def start = threadMX.currentThreadCpuTime c.call() (threadMX.currentThreadCpuTime - start)/1000000
}</lang>
Wall Clock Timing
<lang groovy>def clockRealTime = { Closure c ->
def start = System.currentTimeMillis() c.call() System.currentTimeMillis() - start
}</lang>
Test: <lang groovy>def countTo = { Long n ->
long i = 0; while(i < n) { i += 1L }
}
["CPU time":clockCpuTime, "wall clock time":clockRealTime].each { measurementType, timer ->
println '\n' [100000000L, 1000000000L].each { testSize -> def measuredTime = timer(countTo.curry(testSize)) println "Counting to ${testSize} takes ${measuredTime}ms of ${measurementType}" }
}</lang>
Output:
Counting to 100000000 takes 23150.5484ms of CPU time Counting to 1000000000 takes 233861.0991ms of CPU time Counting to 100000000 takes 24314ms of wall clock time Counting to 1000000000 takes 249005ms of wall clock time
Halon
<lang halon>$t = uptime();
sleep(1);
echo uptime() - $t;</lang>
Haskell
<lang haskell>import System.CPUTime (getCPUTime)
-- We assume the function we are timing is an IO monad computation timeIt :: (Fractional c) => (a -> IO b) -> a -> IO c timeIt action arg = do
startTime <- getCPUTime action arg finishTime <- getCPUTime return $ fromIntegral (finishTime - startTime) / 1000000000000
-- Version for use with evaluating regular non-monadic functions timeIt_ :: (Fractional c) => (a -> b) -> a -> IO c timeIt_ f = timeIt ((`seq` return ()) . f)</lang>
Example
*Main> :m + Text.Printf Data.List *Main Data.List Text.Printf> timeIt' id 4 >>= printf "Identity(4) takes %f seconds.\n" Identity(4) takes 0.0 seconds. *Main Data.List Text.Printf> timeIt' (\x -> foldl' (+) x [1..1000000]) 4 >>= printf "Sum(4) takes %f seconds.\n" Sum(4) takes 0.248015 seconds.
HicEst
<lang HicEst>t_start = TIME() ! returns seconds since midnight SYSTEM(WAIT = 1234) ! wait 1234 milliseconds t_end = TIME()
WRITE(StatusBar) t_end - t_start, " seconds"</lang>
Icon and Unicon
The function 'timef' takes as argument a procedure name and collects performance and timing information including run time (in milliseconds), garbage collection, and memory usage by region.
<lang Icon>procedure timef(f) #: time a function f local gcol,alloc,used,size,runtime,header,x,i
title := ["","total","static","string","block"] # headings collect() # start with collected memory (before baseline) every put(gcol := [], -&collections) # baseline collections count every put(alloc := [], -&allocated) # . total allocated space by region every put(used := [], -&storage) # . currently used space by region - no total every put(size := [], -®ions) # . current size of regions - no total
write("Performance and Timing measurement for ",image(f),":") runtime := &time # base time f() write("Execution time=",&time-runtime," ms.")
every (i := 0, x := &collections) do gcol[i +:= 1] +:= x every (i := 0, x := &allocated ) do alloc[i +:= 1] +:= x every (i := 0, x := &storage ) do used[i +:= 1] +:= x every (i := 0, x := ®ions ) do size[i +:= 1] +:= x
push(gcol,"garbage collections:") push(alloc,"memory allocated:") push(used,"N/A","currently used:") push(size,"N/A","current size:")
write("Memory Region and Garbage Collection Summary (delta):") every (i := 0) <:= *!(title|gcol|alloc|used|size) every x := (title|gcol|alloc|used|size) do {
f := left every writes(f(!x,i + 3)) do f := right write() }
write("Note: static region values should be zero and may not be meaningful.") return end</lang>
Sample usage:<lang Icon>procedure main() timef(perfectnumbers) end
procedure perfectnumbers() ...</lang>
Sample output (from the Perfect Numbers task):
Performance and Timing measurement for procedure perfectnumbers: Perfect numbers from 1 to 10000: 6 28 496 8128 Done. Execution time=416 ms. Memory Region and Garbage Collection Summary (delta): total static string block garbage collections: 2 0 0 2 memory allocated: 1247012 0 24 1246988 currently used: N/A 0 0 248040 current size: N/A 0 0 0 Note: static region values should be zero and may not be meaningful.
Ioke
<lang ioke>use("benchmark")
func = method((1..50000) reduce(+))
Benchmark report(1, 1, func)</lang>
J
Time and space requirements are tested using verbs obtained through the Foreign conjunction (!:
). 6!:2
returns time required for execution, in floating-point measurement of seconds. 7!:2
returns a measurement of space required to execute. Both receive as input a sentence for execution. The verb timespacex
combines these and is available in the standard library.
When the Memoize feature or similar techniques are used, execution time and space can both be affected by prior calculations.
Example
<lang j> (6!:2 , 7!:2) '|: 50 50 50 $ i. 50^3' 0.00488008 3.14829e6
timespacex '|: 50 50 50 $ i. 50^3'
0.00388519 3.14829e6</lang>
Java
<lang java>import java.lang.management.ManagementFactory; import java.lang.management.ThreadMXBean;
public class TimeIt { public static void main(String[] args) { final ThreadMXBean threadMX = ManagementFactory.getThreadMXBean(); assert threadMX.isCurrentThreadCpuTimeSupported(); threadMX.setThreadCpuTimeEnabled(true);
long start, end; start = threadMX.getCurrentThreadCpuTime(); countTo(100000000); end = threadMX.getCurrentThreadCpuTime(); System.out.println("Counting to 100000000 takes "+(end-start)/1000000+"ms"); start = threadMX.getCurrentThreadCpuTime(); countTo(1000000000L); end = threadMX.getCurrentThreadCpuTime(); System.out.println("Counting to 1000000000 takes "+(end-start)/1000000+"ms");
}
public static void countTo(long x){ System.out.println("Counting..."); for(long i=0;i<x;i++); System.out.println("Done!"); } }</lang>
Measures real time rather than CPU time:
<lang java> public static void main(String[] args){ long start, end; start = System.currentTimeMillis(); countTo(100000000); end = System.currentTimeMillis(); System.out.println("Counting to 100000000 takes "+(end-start)+"ms"); start = System.currentTimeMillis(); countTo(1000000000L); end = System.currentTimeMillis(); System.out.println("Counting to 1000000000 takes "+(end-start)+"ms");
}</lang> Output:
Counting... Done! Counting to 100000000 takes 370ms Counting... Done! Counting to 1000000000 takes 3391ms
Julia
<lang julia># v0.6.0
function countto(n::Integer)
i = zero(n) println("Counting...") while i < n i += 1 end println("Done!")
end
@time countto(10 ^ 5) @time countto(10 ^ 10)</lang>
- Output:
Counting... Done! Counting... Done! 0.000109 seconds (15 allocations: 400 bytes) Counting... Done! 0.000127 seconds (15 allocations: 400 bytes)
Kotlin
<lang scala>// version 1.1.2 // need to enable runtime assertions with JVM -ea option
import java.lang.management.ManagementFactory import java.lang.management.ThreadMXBean
fun countTo(x: Int) {
println("Counting..."); (1..x).forEach {} println("Done!")
}
fun main(args: Array<String>) {
val counts = intArrayOf(100_000_000, 1_000_000_000) val threadMX = ManagementFactory.getThreadMXBean() assert(threadMX.isCurrentThreadCpuTimeSupported) threadMX.isThreadCpuTimeEnabled = true for (count in counts) { val start = threadMX.currentThreadCpuTime countTo(count) val end = threadMX.currentThreadCpuTime println("Counting to $count takes ${(end-start)/1000000}ms") }
}</lang> This is a typical result (sometimes the second figure is only about 1400ms - no idea why)
- Output:
Counting... Done! Counting to 100000000 takes 179ms Counting... Done! Counting to 1000000000 takes 3527ms
Lasso
<lang Lasso>local(start = micros) loop(100000) => { 'nothing is outout because no autocollect' } 'time for 100,000 loop repititions: '+(micros - #start)+' microseconds'</lang>
Lingo
<lang lingo>on testFunc ()
repeat with i = 1 to 1000000 x = sqrt(log(i)) end repeat
end</lang>
<lang lingo>ms = _system.milliseconds testFunc() ms = _system.milliseconds - ms put "Execution time in ms:" && ms -- "Execution time in ms: 983"</lang>
Logo
on a Unix system
This is not an ideal method; Logo does not expose a timer (except for the WAIT command) so we use the Unix "date" command to get a second timer.
<lang logo>to time
output first first shell "|date +%s|
end to elapsed :block
localmake "start time run :block (print time - :start [seconds elapsed])
end
elapsed [wait 300] ; 5 seconds elapsed</lang>
Lua
<lang lua>function Test_Function()
for i = 1, 10000000 do local s = math.log( i ) s = math.sqrt( s ) end
end
t1 = os.clock()
Test_Function()
t2 = os.clock()
print( os.difftime( t2, t1 ) )</lang>
M2000 Interpreter
We use Profiler to reset timer, and Timecount to read time in milliseconds as a double, with nanoseconds for resolution. Internal use of QueryPerformanceCounter from Windows Api. In this example we get times for use of same module with different variable types. sum=limit-limit make sum 0 to the same type of limit,and using n=sum and n++ we make n=1 using same type as sum.
10000% is Integer 16bit
10000& is Long 32bit
10000@ is Decimal
10000# is Currency
10000~ is Float
10000 is Double (default)
<lang M2000 Interpreter> Module Checkit {
Module sumtolimit (limit) { sum=limit-limit n=sum n++ while limit {sum+=limit*n:limit--:n-!} } Cls ' clear screen Profiler sumtolimit 10000% Print TimeCount Profiler sumtolimit 10000& Print TimeCount Profiler sumtolimit 10000# Print TimeCount Profiler sumtolimit 10000@ Print TimeCount Profiler sumtolimit 10000~ Print TimeCount Profiler sumtolimit 10000 Print TimeCount
} Checkit </lang>
Maple
The built-in command CodeTools:-Usage can compute the "real" time for the length of the computation or the "cpu" time for the computation. The following examples find the real time and cpu time for computing the integer factors for 32!+1. <lang maple>CodeTools:-Usage(ifactor(32!+1), output = realtime, quiet);</lang> <lang maple>CodeTools:-Usage(ifactor(32!+1), output = cputime, quiet);</lang>
Mathematica
<lang Mathematica>AbsoluteTiming[x];</lang> where x is an operation. Example calculating a million digits of Sqrt[3]: <lang Mathematica>AbsoluteTiming[N[Sqrt[3], 10^6]]</lang> gives: <lang Mathematica>{0.000657, 1.7320508075688772935274463......}</lang> First elements if the time in seconds, second elements if the result from the operation. Note that I truncated the result.
Maxima
<lang maxima>f(n) := if n < 2 then n else f(n - 1) + f(n - 2)$
/* First solution, call the time function with an output line number, it gives the time taken to compute that line.
Here it's assumed to be %o2 */
f(24); 46368
time(%o2); [0.99]
/* Second solution, change a system flag to print timings for all following lines */ showtime: true$
f(24); Evaluation took 0.9400 seconds (0.9400 elapsed) 46368</lang>
Nim
<lang nim>import times, os
proc doWork(x) =
var n = x for i in 0..10000000: n += i echo n
template time(s: stmt): expr =
let t0 = cpuTime() s cpuTime() - t0
echo time(doWork(100))</lang> Output:
2.2000000000000000e-01
OCaml
<lang ocaml>let time_it action arg =
let start_time = Sys.time () in ignore (action arg); let finish_time = Sys.time () in finish_time -. start_time</lang>
Example
# Printf.printf "Identity(4) takes %f seconds.\n" (time_it (fun x -> x) 4);; Identity(4) takes 0.000000 seconds. - : unit = () # let sum x = let num = ref x in for i = 0 to 999999 do num := !num + i done; !num;; val sum : int -> int = <fun> # Printf.printf "Sum(4) takes %f seconds.\n" (time_it sum 4);; Sum(4) takes 0.084005 seconds. - : unit = ()
Oforth
bench allows to calculate how long a runnable takes to execute.
Result is microseconds.
It uses difference between system time, not processing time.
- Output:
>#[ 0 1000 seq apply(#+) ] bench . 267 500500 ok
Oz
<lang oz>declare
%% returns milliseconds fun {TimeIt Proc} Before = {Now} in {Proc} {Now} - Before end
fun {Now} {Property.get 'time.total'} end
in
{Show {TimeIt proc {$} {FoldL {List.number 1 1000000 1} Number.'+' 4 _} end} }</lang>
PARI/GP
This version, by default, returns just the CPU time used by gp, not the delta of wall times. PARI can be compiled to use wall time if you prefer: configure with --time=ftime
instead of --time=
getrusage
, --time=clock_gettime
, or --time=times
. See Appendix A, section 2.2 of the User's Guide to PARI/GP.
<lang parigp>time(foo)={
foo(); gettime();
}</lang>
Alternate version:
<lang parigp>time(foo)={
my(start=getabstime()); foo(); getabstime()-start;
}</lang>
Perl
Example of using the built-in Benchmark core module - it compares two versions of recursive factorial functions: <lang perl>use Benchmark; use Memoize;
sub fac1 {
my $n = shift; return $n == 0 ? 1 : $n * fac1($n - 1);
} sub fac2 {
my $n = shift; return $n == 0 ? 1 : $n * fac2($n - 1);
} memoize('fac2');
my $result = timethese(100000, {
'fac1' => sub { fac1(50) }, 'fac2' => sub { fac2(50) },
}); Benchmark::cmpthese($result);</lang> Output:
Benchmark: timing 100000 iterations of fac1, fac2... fac1: 6 wallclock secs ( 5.45 usr + 0.00 sys = 5.45 CPU) @ 18348.62/s (n=100000) fac2: 1 wallclock secs ( 0.84 usr + 0.00 sys = 0.84 CPU) @ 119047.62/s (n=100000) Rate fac1 fac2 fac1 18349/s -- -85% fac2 119048/s 549% --
Example without using Benchmark: <lang perl>sub cpu_time {
my ($user,$system,$cuser,$csystem) = times; $user + $system
}
sub time_it {
my $action = shift; my $startTime = cpu_time(); $action->(@_); my $finishTime = cpu_time(); $finishTime - $startTime
}
printf "Identity(4) takes %f seconds.\n", time_it(sub {@_}, 4);
- outputs "Identity(4) takes 0.000000 seconds."
sub sum {
my $x = shift; foreach (0 .. 999999) { $x += $_; } $x
}
printf "Sum(4) takes %f seconds.\n", time_it(\&sum, 4);
- outputs "Sum(4) takes 0.280000 seconds."</lang>
Perl 6
Follows modern trend toward measuring wall-clock time, since CPU time is becoming rather ill-defined in the age of multicore, and doesn't reflect IO overhead in any case. <lang perl6>my $start = now; (^100000).pick(1000); say now - $start;</lang>
- Output:
0.02301709
Phix
Measures wall-clock time. On Windows the resolution is about 15ms. <lang Phix>atom t0 = time() some_procedure() printf(1,"%3.2fs.\n",time()-t0)</lang>
PicoLisp
There is a built-in function 'bench' for that. However, it measures wall-clock time, because for practical purposes the real time needed by a task (including I/O and communication) is more meaningful. There is another function, 'tick', which also measures user time, and is used by the profiling tools. <lang PicoLisp>: (bench (do 1000000 (* 3 4))) 0.080 sec -> 12</lang>
PL/I
<lang PL/I>declare (start_time, finish_time) float (18);
start_time = secs();
do i = 1 to 10000000;
/* something to be repeated goes here. */
end; finish_time = secs();
put skip edit ('elapsed time=', finish_time - start_time, ' seconds')
(A, F(10,3), A); /* gives the result to thousandths of a second. */
/* Note: using the SECS function takes into account the clock */ /* going past midnight. */</lang>
PowerShell
<lang PowerShell> function fun($n){
$res = 0 if($n -gt 0) { 1..$n | foreach{ $a, $b = $_, ($n+$_) $res += $a + $b }
} $res
} "$((Measure-Command {fun 10000}).TotalSeconds) Seconds" </lang> Output:
0.820712 Seconds
PureBasic
Built in timer
This version uses the built in timer, on Windows it has an accuracy of ~10-15 msec. <lang Purebasic>Procedure Foo(Limit)
Protected i, palindromic, String$ For i=0 To Limit String$=Str(i) If String$=ReverseString(String$) palindromic+1 EndIf Next ProcedureReturn palindromic
EndProcedure
If OpenConsole()
Define Start, Stop, cnt PrintN("Starting timing of a calculation,") PrintN("for this we test how many of 0-1000000 are palindromic.") Start=ElapsedMilliseconds() cnt=Foo(1000000) Stop=ElapsedMilliseconds() PrintN("The function need "+Str(stop-Start)+" msec,") PrintN("and "+Str(cnt)+" are palindromic.") Print("Press ENTER to exit."): Input()
EndIf</lang>
Starting timing of a calculation, for this we test how many of 0-1000000 are palindromic. The function need 577 msec, and 1999 are palindromic. Press ENTER to exit.
Hi-res version
This version uses a hi-res timer, but it is Windows only. <lang PureBasic>If OpenConsole()
Define Timed.f, cnt PrintN("Starting timing of a calculation,") PrintN("for this we test how many of 0-1000000 are palindromic.") ; Dependent on Droopy-library If MeasureHiResIntervalStart() ; Same Foo() as above... cnt=Foo(1000000) Timed=MeasureHiResIntervalStop() EndIf PrintN("The function need "+StrF(Timed*1000,3)+" msec,") PrintN("and "+Str(cnt)+" are palindromic.") Print("Press ENTER to exit."): Input()
EndIf</lang>
Starting timing of a calculation, for this we test how many of 0-1000000 are palindromic. The function need 604.341 msec, and 1999 are palindromic. Press ENTER to exit.
This version still relies on the Windows API but does not make use of any additional libraries. <lang PureBasic>Procedure.f ticksHQ(reportIfPresent = #False)
Static maxfreq.q Protected T.q If reportIfPresent Or maxfreq = 0 QueryPerformanceFrequency_(@maxfreq) If maxfreq ProcedureReturn 1.0 Else ProcedureReturn 0 EndIf EndIf QueryPerformanceCounter_(@T) ProcedureReturn T / maxfreq ;Result is in milliseconds
EndProcedure
If OpenConsole()
Define timed.f, cnt PrintN("Starting timing of a calculation,") PrintN("for this we test how many of 0-1000000 are palindromic.") ; Dependent on Windows API If ticksHQ(#True) timed = ticksHQ() ;start time ; Same Foo() as above... cnt = Foo(1000000) timed = ticksHQ() - timed ;difference EndIf PrintN("The function need " + StrF(timed * 1000, 3) + " msec,") PrintN("and " + Str(cnt) + " are palindromic.") Print("Press ENTER to exit."): Input()
EndIf</lang>
Sample output:
Starting timing of a calculation, for this we test how many of 0-1000000 are palindromic. The function need 174.811 msec, and 1999 are palindromic.
Python
Given function and arguments return a time (in microseconds) it takes to make the call.
Note: There is an overhead in executing a function that does nothing. <lang python>import sys, timeit def usec(function, arguments):
modname, funcname = __name__, function.__name__ timer = timeit.Timer(stmt='%(funcname)s(*args)' % vars(), setup='from %(modname)s import %(funcname)s; args=%(arguments)r' % vars()) try: t, N = 0, 1 while t < 0.2: t = min(timer.repeat(repeat=3, number=N)) N *= 10 microseconds = round(10000000 * t / N, 1) # per loop return microseconds except: timer.print_exc(file=sys.stderr) raise
def nothing(): pass def identity(x): return x</lang>
Example
>>> print usec(nothing, []) 1.7 >>> print usec(identity, [1]) 2.2 >>> print usec(pow, (2, 100)) 3.3 >>> print [usec(qsort, (range(n),)) for n in range(10)] [2.7, 2.8, 31.4, 38.1, 58.0, 76.2, 100.5, 130.0, 149.3, 180.0]
using qsort() from Quicksort. Timings show that the implementation of qsort() has quadratic dependence on sequence length N for already sorted sequences (instead of O(N*log(N)) in average).
R
R has a built-in function, system.time, to calculate this. <lang R># A task foo <- function() {
for(i in 1:10) { mat <- matrix(rnorm(1e6), nrow=1e3) mat^-0.5 }
}
- Time the task
timer <- system.time(foo())
- Extract the processing time
timer["user.self"]</lang> For a breakdown of processing time by function, there is Rprof. <lang R>Rprof() foo() Rprof(NULL) summaryRprof()</lang>
Racket
<lang racket>
- lang racket
(define (fact n) (if (zero? n) 1 (* n (fact (sub1 n))))) (time (fact 5000)) </lang>
Raven
<lang Raven>define doId use $x
$x dup * $x /
define doPower use $v, $p
$v $p pow
define doSort
group 20000 each choose list sort reverse
define timeFunc use $fName
time as $t1 $fName "" prefer call time as $t2 $fName $t2 $t1 -"%.4g secs for %s\n" print
"NULL" timeFunc 42 "doId" timeFunc 12 2 "doPower" timeFunc "doSort" timeFunc</lang>
- Output:
2.193e-05 secs for NULL 2.003e-05 secs for doId 4.601e-05 secs for doPower 3.028 secs for doSort
Retro
Retro has a time function returning the current time in seconds. This can be used to build a simple timing function:
<lang Retro>: .runtime ( a- ) time [ do time ] dip - "\n%d\n" puts ;
- test 20000 [ putn space ] iterd ;
&test .runtime</lang>
Finer measurements are not possible with the standard implementation.
REXX
elapsed time version
REXX doesn't have a language feature for obtaining true CPU time (except under
IBM mainframes which have commands that can retrieve such times), but it does
have a built-in function for elapsed time(s).
The main reason for the true CPU time omission is that REXX was developed under VM/CMS and
there's a way to easily query the host (VM/CP) to indicate how much true CPU time was used by
(normally) your own userID. The result can then be placed into a REXX variable (as an option).
<lang rexx>/*REXX program displays the elapsed time for a REXX function (or subroutine). */
arg reps . /*obtain an optional argument from C.L.*/
if reps== then reps=100000 /*Not specified? No, then use default.*/
call time 'Reset' /*only the 1st character is examined. */
junk = silly(reps) /*invoke the SILLY function (below). */
/*───► CALL SILLY REPS also works.*/
/* The E is for elapsed time.*/ /* │ ─ */ /* ┌────◄───┘ */ /* │ */ /* ↓ */
say 'function SILLY took' format(time("E"),,2) 'seconds for' reps "iterations."
/* ↑ */ /* │ */ /* ┌────────►───────┘ */ /* │ */ /* The above 2 for the FORMAT function displays the time with*/ /* two decimal digits (rounded) past the decimal point). Using */ /* a 0 (zero) would round the time to whole seconds. */
exit /*stick a fork in it, we're all done. */ /*────────────────────────────────────────────────────────────────────────────*/ silly: procedure /*chew up some CPU time doing some silly stuff.*/
do j=1 for arg(1) /*wash, apply, lather, rinse, repeat. ··· */ @.j=random() date() time() digits() fuzz() form() xrange() queued() end /*j*/ return j-1</lang>
output when using a personal computer built in the 20th century:
function SILLY took 3.54 seconds for 100000 iterations.
output when using a personal computer built in the 21st century:
function SILLY took 0.44 seconds for 100000 iterations.
output when using an IBM mainframe with MVS/TSO:
function SILLY took 0.69 seconds for 100000 iterations.
CPU time used version
This version only works with Regina REXX as the J option (for the time BIF) is a Regina extension.
Since the silly function (by far) consumes the bulk of the CPU time of the REXX program, what is
being measured is essentially the same as the wall clock time (duration) of the function execution; the
overhead of the invocation is minimal compared to the overall time used.
<lang rexx>/*REXX program displays the elapsed time for a REXX function (or subroutine). */
arg reps . /*obtain an optional argument from C.L.*/
if reps== then reps=100000 /*Not specified? No, then use default.*/
call time 'Reset' /*only the 1st character is examined. */
junk = silly(reps) /*invoke the SILLY function (below). */
/*───► CALL SILLY REPS also works.*/
/* The J is for the CPU time used */ /* │ by the REXX program since */ /* ┌───────┘ since the time was RESET. */ /* │ This is a Regina extension.*/ /* ↓ */
say 'function SILLY took' format(time("J"),,2) 'seconds for' reps "iterations."
/* ↑ */ /* │ */ /* ┌────────►───────┘ */ /* │ */ /* The above 2 for the FORMAT function displays the time with*/ /* two decimal digits (rounded) past the decimal point). Using */ /* a 0 (zero) would round the time to whole seconds. */
exit /*stick a fork in it, we're all done. */ /*────────────────────────────────────────────────────────────────────────────*/ silly: procedure /*chew up some CPU time doing some silly stuff.*/
do j=1 for arg(1) /*wash, apply, lather, rinse, repeat. ··· */ @.j=random() date() time() digits() fuzz() form() xrange() queued() end /*j*/ return j-1</lang>
output is essentially identical to the previous examples.
Ring
<lang ring> beginTime = TimeList()[13] for n = 1 to 10000000
n = n + 1
next endTime = TimeList()[13] elapsedTime = endTime - beginTime see "Elapsed time = " + elapsedTime + nl </lang>
Ruby
Ruby's Benchmark module provides a way to generate nice reports (numbers are in seconds): <lang ruby>require 'benchmark'
Benchmark.bm(8) do |x|
x.report("nothing:") { } x.report("sum:") { (1..1_000_000).inject(4) {|sum, x| sum + x} }
end</lang> Output:
user system total real nothing: 0.000000 0.000000 0.000000 ( 0.000014) sum: 2.700000 0.400000 3.100000 ( 3.258348)
You can get the total time as a number for later processing like this: <lang ruby>Benchmark.measure { whatever }.total</lang>
Scala
Define a time
function that returns the elapsed time (in ms) to execute a block of code.
<lang scala>
def time(f: => Unit)={
val s = System.currentTimeMillis
f
System.currentTimeMillis - s
}
</lang>
Can be called with a code block:
<lang scala>
println(time {
for(i <- 1 to 10000000) {}
})
</lang>
Or with a function:
<lang scala>
def count(i:Int) = for(j <- 1 to i){}
println(time (count(10000000))) </lang>
Scheme
<lang scheme>(time (some-function))</lang>
Seed7
<lang seed7>$ include "seed7_05.s7i";
include "time.s7i"; include "duration.s7i";
const func integer: identity (in integer: x) is
return x;
const func integer: sum (in integer: num) is func
result var integer: result is 0; local var integer: number is 0; begin result := num; for number range 1 to 1000000 do result +:= number; end for; end func;
const func duration: timeIt (ref func integer: aFunction) is func
result var duration: result is duration.value; local var time: before is time.value; begin before := time(NOW); ignore(aFunction); result := time(NOW) - before; end func;
const proc: main is func
begin writeln("Identity(4) takes " <& timeIt(identity(4))); writeln("Sum(4) takes " <& timeIt(sum(4))); end func;</lang>
- Output:
of interpreted program
Identity(4) takes 0-00-00 00:00:00.000163 Sum(4) takes 0-00-00 00:00:00.131823
- Output:
of compiled program (optimized with -O2)
Identity(4) takes 0-00-00 00:00:00.000072 Sum(4) takes 0-00-00 00:00:00.000857
Sidef
<lang ruby>var benchmark = frequire('Benchmark')
func fac_rec(n) {
n == 0 ? 1 : (n * __FUNC__(n - 1))
}
func fac_iter(n) {
var prod = 1 n.times { |i| prod *= i } prod
}
var result = benchmark.timethese(-3, Hash(
'fac_rec' => { fac_rec(20) }, 'fac_iter' => { fac_iter(20) },
))
benchmark.cmpthese(result)</lang>
- Output:
Benchmark: running fac_iter, fac_rec for at least 3 CPU seconds... fac_iter: 3 wallclock secs ( 3.23 usr + 0.00 sys = 3.23 CPU) @ 7331.89/s (n=23682) fac_rec: 3 wallclock secs ( 3.19 usr + 0.00 sys = 3.19 CPU) @ 3551.72/s (n=11330) Rate fac_rec fac_iter fac_rec 3552/s -- -52% fac_iter 7332/s 106% --
Slate
<lang slate> [inform: 2000 factorial] timeToRun. </lang>
Smalltalk
(Squeak/Pharo) <lang smalltalk> Time millisecondsToRun: [ Transcript show: 2000 factorial ]. </lang>
Standard ML
<lang sml>fun time_it (action, arg) = let
val timer = Timer.startCPUTimer () val _ = action arg val times = Timer.checkCPUTimer timer
in
Time.+ (#usr times, #sys times)
end</lang>
Example
- print ("Identity(4) takes " ^ Time.toString (time_it (fn x => x, 4)) ^ " seconds.\n"); Identity(4) takes 0.000 seconds. val it = () : unit - fun sum (x:IntInf.int) = let fun loop (i, sum) = if i >= 1000000 then sum else loop (i + 1, sum + i) in loop (0, x) end; val sum = fn : IntInf.int -> IntInf.int - print ("Sum(4) takes " ^ Time.toString (time_it (sum, 4)) ^ " seconds.\n"); Sum(4) takes 0.220 seconds. val it = () : unit
Stata
Stata can track up to 100 timers. See timer in Stata help.
<lang stata>program timer_test timer clear 1 timer on 1 sleep `0' timer off 1 timer list 1 end
. timer_test 1000
1: 1.01 / 1 = 1.0140</lang>
Tcl
The Tcl time
command returns the real time elapsed
averaged over a number of iterations.
<lang tcl>proc sum_n {n} {
for {set i 1; set sum 0.0} {$i <= $n} {incr i} {set sum [expr {$sum + $i}]} return [expr {wide($sum)}]
}
puts [time {sum_n 1e6} 100] puts [time {} 100]</lang>
- Output:
163551.0 microseconds per iteration 0.2 microseconds per iteration
TorqueScript
Greek2me 02:16, 19 June 2012 (UTC)
Returns average time elapsed from many iterations. <lang TorqueScript> function benchmark(%times,%function,%a,%b,%c,%d,%e,%f,%g,%h,%i,%j,%k,%l,%m,%n,%o) { if(!isFunction(%function)) { warn("BENCHMARKING RESULT FOR" SPC %function @ ":" NL "Function does not exist."); return -1; }
%start = getRealTime();
for(%i=0; %i < %times; %i++) { call(%function,%a,%b,%c,%d,%e,%f,%g,%h,%i,%j,%k,%l,%m,%n,%o); }
%end = getRealTime();
%result = (%end-%start) / %times;
warn("BENCHMARKING RESULT FOR" SPC %function @ ":" NL %result);
return %result; } </lang>
- Example:
<lang TorqueScript> function exampleFunction(%var1,%var2) { //put stuff here }
benchmark(500,"exampleFunction","blah","variables");
==> BENCHMARKING RESULT FOR exampleFunction: ==> 13.257 </lang>
TUSCRIPT
<lang tuscript> $$ MODE TUSCRIPT SECTION test LOOP n=1,999999 rest=MOD (n,1000) IF (rest==0) Print n ENDLOOP ENDSECTION time_beg=TIME () DO test time_end=TIME () interval=TIME_INTERVAL (seconds,time_beg,time_end) PRINT "'test' start at ",time_beg PRINT "'test' ends at ",time_end PRINT "'test' takes ",interval," seconds" </lang>
- Output:
'test' start at 2011-01-15 14:38:22 'test' ends at 2011-01-15 14:38:31 'test' takes 9 seconds
UNIX Shell
<lang bash>$ time sleep 1</lang>
real 0m1.074s user 0m0.001s sys 0m0.006s
VBA
<lang vb>Public Declare Function GetTickCount Lib "kernel32.dll" () As Long Private Function identity(x As Long) As Long
For j = 0 To 1000 identity = x Next j
End Function Private Function sum(ByVal num As Long) As Long
Dim t As Long For j = 0 To 1000 t = num For i = 0 To 10000 t = t + i Next i Next j sum = t
End Function Private Sub time_it()
Dim start_time As Long, finis_time As Long start_time = GetTickCount identity 1 finis_time = GetTickCount Debug.Print "1000 times Identity(1) takes "; (finis_time - start_time); " milliseconds" start_time = GetTickCount sum 1 finis_time = GetTickCount Debug.Print "1000 times Sum(1) takes "; (finis_time - start_time); " milliseconds"
End Sub</lang>
- Output:
1000 times Identity(1) takes 0 seconds 1000 times Sum(1) takes 296 seconds
Wart
<lang python>time 1+1 30000/1000000 # in microseconds => 2</lang>
XPL0
This works fine under pure DOS but has problems under Windows. Windows can execute other processes, although it could be argued that this should be included as part of the total time to accomplish the task at hand. DOS does go off to service a timer interrupt, but it's usually very fast, although beware of TSRs that hook this interrupt handler.
There's a more serious problem with the GetTime intrinsic under Windows XP. GetTime provides microsecond resolution by combining the BIOS timer interrupt count at location 046C with the count in the 8254 chip (or its equivalent). Unfortunately, Windows virtualizes the 8254 and thus the two values can be out of sync.
<lang XPL0>include c:\cxpl\codes; int T0, T1, I; [T0:= GetTime; for I:= 1, 1_000_000 do []; T1:= GetTime; IntOut(0, T1-T0); Text(0, " microseconds^M^J"); ]</lang>
- Example output:
for a Duron 850 running DOS 5.0
2354 microseconds
Yabasic
<lang Yabasic>sub count(n) local i
for i = 1 to n next i end sub
count(1000000)
print peek("millisrunning"), " milliseconds"
t0 = peek("millisrunning") count(10000000) print peek("millisrunning")-t0, " milliseconds"</lang>
zkl
In order to be as OS independent as possible, only system time is available. <lang zkl>t:=Time.Clock.time; Atomic.sleep(3); (Time.Clock.time - t).println();</lang>
- Output:
3
- Programming Tasks
- Programming environment operations
- Date and time
- Batch File/Omit
- GUISS/Omit
- ML/I/Omit
- 8051 Assembly
- ACL2
- Ada
- Aime
- AutoHotkey
- BaCon
- BASIC
- Batch File
- BBC BASIC
- Bracmat
- C
- C++
- C sharp
- Clojure
- Common Lisp
- D
- E
- Elena
- Elixir
- Erlang
- Euphoria
- F Sharp
- Factor
- Forth
- Fortran
- FreeBASIC
- GAP
- Go
- Groovy
- Halon
- Haskell
- HicEst
- Icon
- Unicon
- Ioke
- J
- Java
- Julia
- Kotlin
- Lasso
- Lingo
- Logo
- Lua
- M2000 Interpreter
- Maple
- Mathematica
- Maxima
- Nim
- OCaml
- Oforth
- Oz
- PARI/GP
- Perl
- Perl 6
- Phix
- PicoLisp
- PL/I
- PowerShell
- PureBasic
- Droopy
- Python
- R
- Racket
- Raven
- Retro
- REXX
- Ring
- Ruby
- Scala
- Scheme
- Seed7
- Sidef
- Slate
- Smalltalk
- Standard ML
- Stata
- Tcl
- TorqueScript
- TUSCRIPT
- UNIX Shell
- VBA
- Wart
- XPL0
- Yabasic
- Zkl