Solve a Rubik's cube: Difference between revisions

Line 23:

For the single line example, typical timings are around 240 milliseconds which is much faster than Kotlin due, no doubt, to JVM warm up time.

<~~lang~~ go>/**********************************************************************

<syntaxhighlight lang="go">/**********************************************************************

*

* A cube 'state' is an int array with 40 entries, the first 20

Line 310:

fmt.Println("\nAverage number of moves =", float64(aggregateMoves)/float64(lineCount))

fmt.Println("\nAverage time =", elapsedTime/int64(lineCount), "milliseconds")

}</~~lang~~>

}</syntaxhighlight>

Line 320:

=={{header|Julia}}==

<~~lang~~ julia>#=**********************************************************************

<syntaxhighlight lang="julia">#=**********************************************************************

*

* A cube 'state' is a vector<int> with 40 entries, the first 20

Line 511:

pochmann("rubikdata.txt")

</~~lang~~>{{out}}

</syntaxhighlight>{{out}}

Using the 100-line database:

Line 627:

To aid readability I've also inserted spaces between each move in the results and added the total moves needed for each line.

<~~lang~~ scala>// version 1.2.21

<syntaxhighlight lang="scala">// version 1.2.21

/**********************************************************************

Line 859:

println("\nAverage number of moves = ${aggregateMoves.toDouble() / lineCount}")

println("\nAverage time = ${elapsedTime / lineCount} milliseconds")

}</~~lang~~>

}</syntaxhighlight>

Line 987:

=={{header|Nim}}==

<syntaxhighlight lang="nim">#[

<lang Nim>#[

**********************************************************************

*

Line 1,190:

let elapsedTime = cpuTime() - startTime

echo &"\nAverage number of moves = {aggregateMoves / lineCount}"

echo &"\nAverage time = {elapsedTime * 1000 / lineCount.toFloat:.2f} milliseconds"</~~lang~~>

echo &"\nAverage time = {elapsedTime * 1000 / lineCount.toFloat:.2f} milliseconds"</syntaxhighlight>

Line 1,319:

The final stage (pll) would probably benefit the most from being replaced with standard algorithms.

While technically this works under pwa/p2js, you should expect a blank screen for nearly 3 minutes.

--

-- demo\rosetta\rubik_cfop.exw

Line 1,708:

end procedure

main()

The "hardest case" from http://www.cube20.org with a high threshold. You can try this manually.

Line 1,774:

Translation/de-golf(hrumph) of Tomas Sirgedas' winning entry from http://tomas.rokicki.com/cubecontest as held in 2004.

Faster and shorter solutions (in most cases) than cfop, however probably nigh on impossible to debug/enhance...

--

-- demo\rosetta\rubik_tomas.exw

Line 1,990:

printf(1,"%s (%d move%s)\n",tomas("UL DL RF UB FD BR DB UF DR UR BL FL FDR BLU DLB URB RUF FLD BRD FUL"))

<pre>

Line 2,003:

Despite this, the script is taking an average of just over a second to calculate the number of moves for each line which is probably not too bad for an interpreted language.

<~~lang~~ ecmascript>/**********************************************************************

<syntaxhighlight lang="ecmascript">/**********************************************************************

*

* A cube 'state' is an int array with 40 entries, the first 20

Line 2,260:

var elapsedTime = ((endTime - startTime) * 1000).round

System.print("\nAverage number of moves = %(aggregateMoves/lineCount)")

System.print("\nAverage time = %(elapsedTime/lineCount) milliseconds")</~~lang~~>

System.print("\nAverage time = %(elapsedTime/lineCount) milliseconds")</syntaxhighlight>