Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.javascript > #34224 > unrolled thread

Sort at least 5 times faster then native and quicksort

Back to article view | Back to comp.lang.javascript

Contents

  Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-10-28 21:27 -0700
    Re: Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-10-28 21:41 -0700
    Re: Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-10-29 03:53 -0700
      Re: Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-10-29 04:33 -0700
    Re: Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-10-29 04:37 -0700
    Re: Sort at least 5 times faster then native and quicksort Ben Bacarisse <ben.usenet@bsb.me.uk> - 2017-10-29 12:10 +0000
      Re: Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-10-29 05:53 -0700
        Re: Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-10-29 06:01 -0700
        Re: Sort at least 5 times faster then native and quicksort Ben Bacarisse <ben.usenet@bsb.me.uk> - 2017-10-29 14:47 +0000
          Re: Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-10-29 08:13 -0700
            Re: Sort at least 5 times faster then native and quicksort Ben Bacarisse <ben.usenet@bsb.me.uk> - 2017-10-29 15:46 +0000
              Re: Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-10-29 08:49 -0700
                Re: Sort at least 5 times faster then native and quicksort Tim Streater <timstreater@greenbee.net> - 2017-10-29 16:16 +0100
              Re: Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-10-29 08:52 -0700
              Re: Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-10-29 08:58 -0700
                Re: Sort at least 5 times faster then native and quicksort Ben Bacarisse <ben.usenet@bsb.me.uk> - 2017-10-29 16:18 +0000
                  Re: Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-10-29 09:36 -0700
                  Re: Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-10-29 09:48 -0700
                    Re: Sort at least 5 times faster then native and quicksort Ben Bacarisse <ben.usenet@bsb.me.uk> - 2017-10-29 18:02 +0000
                      Re: Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-10-29 20:51 -0700
                        Re: Sort at least 5 times faster then native and quicksort Ben Bacarisse <ben.usenet@bsb.me.uk> - 2017-10-30 09:31 +0000
                          Re: Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-10-30 03:14 -0700
                            Re: Sort at least 5 times faster then native and quicksort Ben Bacarisse <ben.usenet@bsb.me.uk> - 2017-10-30 10:31 +0000
                              Re: Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-10-30 03:35 -0700
                                Re: Sort at least 5 times faster then native and quicksort Ben Bacarisse <ben.usenet@bsb.me.uk> - 2017-10-30 14:19 +0000
                                  Re: Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-11-02 14:38 -0700
                                  Re: Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-11-02 14:42 -0700
                              Re: Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-10-30 03:44 -0700
                              Re: Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-10-30 03:50 -0700
                            Re: Sort at least 5 times faster then native and quicksort Wally W. <ww84wa@aim.com> - 2017-10-30 06:40 -0400
                              Re: Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-10-30 03:48 -0700
                          Re: Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-10-30 03:32 -0700
                            Re: Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-10-30 03:37 -0700
          Re: Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-10-29 08:35 -0700
      Re: Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-10-29 06:07 -0700
        Re: Sort at least 5 times faster then native and quicksort Ben Bacarisse <ben.usenet@bsb.me.uk> - 2017-10-29 14:37 +0000
      Re: Sort at least 5 times faster then native and quicksort jonas.thornvall@gmail.com - 2017-10-29 06:23 -0700
        Re: Sort at least 5 times faster then native and quicksort Ben Bacarisse <ben.usenet@bsb.me.uk> - 2017-10-29 14:50 +0000

Page 1 of 2  [1] 2  Next page →

#34224 — Sort at least 5 times faster then native and quicksort

My sort Algorithm is at least 5 time faster then native and quicks sort for densely populated samples "where numbers have 1 or more occurences".


But usual there is some weird aspect i can't find out what i do wrong, if you set startrange for the random numbergererator higher then 0 values will turn up undefined i don't get it.

But the timings are correct, 5 times faster for 100k of integers, and best of all the algorithm is linear, but it purely for sorting numbers.

You guys probably see right away what goes wrong, when you change the start range. I do not.

http://anybase.co.nf/jtsort.html

[toc] | [next] | [standalone]

#34225

Den söndag 29 oktober 2017 kl. 05:27:47 UTC+1 skrev jonas.t...@gmail.com:
> My sort Algorithm is at least 5 time faster then native and quicks sort for densely populated samples "where numbers have 1 or more occurences".
> 
> 
> But usual there is some weird aspect i can't find out what i do wrong, if you set startrange for the random numbergererator higher then 0 values will turn up undefined i don't get it.
> 
> But the timings are correct, 5 times faster for 100k of integers, and best of all the algorithm is linear, but it purely for sorting numbers.
> 
> You guys probably see right away what goes wrong, when you change the start range. I do not.
> 
> http://anybase.co.nf/jtsort.html

If you can see ways to make it faster i would like to hear.

[toc] | [prev] | [next] | [standalone]

#34227

Den söndag 29 oktober 2017 kl. 05:27:47 UTC+1 skrev jonas.t...@gmail.com:
> My sort Algorithm is at least 5 time faster then native and quicks sort for densely populated samples "where numbers have 1 or more occurences".
> 
> 
> But usual there is some weird aspect i can't find out what i do wrong, if you set startrange for the random numbergererator higher then 0 values will turn up undefined i don't get it.
> 
> But the timings are correct, 5 times faster for 100k of integers, and best of all the algorithm is linear, but it purely for sorting numbers.
> 
> You guys probably see right away what goes wrong, when you change the start range. I do not.
> 
> http://anybase.co.nf/jtsort.html

For 1 million values jtsort take 543 ms javascript 3948 ms so it is 8 time faster.

For 10 million values jtsort take 30474 ms "30 sec", native sort more then 207047 "3 minutes 27 sec".

[toc] | [prev] | [next] | [standalone]

#34228

Den söndag 29 oktober 2017 kl. 11:54:00 UTC+1 skrev jonas.t...@gmail.com:
> Den söndag 29 oktober 2017 kl. 05:27:47 UTC+1 skrev jonas.t...@gmail.com:
> > My sort Algorithm is at least 5 time faster then native and quicks sort for densely populated samples "where numbers have 1 or more occurences".
> > 
> > 
> > But usual there is some weird aspect i can't find out what i do wrong, if you set startrange for the random numbergererator higher then 0 values will turn up undefined i don't get it.
> > 
> > But the timings are correct, 5 times faster for 100k of integers, and best of all the algorithm is linear, but it purely for sorting numbers.
> > 
> > You guys probably see right away what goes wrong, when you change the start range. I do not.
> > 
> > http://anybase.co.nf/jtsort.html
> 
> For 1 million values jtsort take 543 ms javascript 3948 ms so it is 8 time faster.
> 
> For 10 million values jtsort take 30474 ms "30 sec", native sort more then 207047 "3 minutes 27 sec".

For densely populated samples at least 5 times faster then javascript native sort and quicksort. Well it is just to sort numbers, but there is other approaches to sort strings, that are multiple times faster then quicksort and javascript native sort.
When script first load it takes time 5-10 sec because i generate and printout 99000 numbers to a form, if you chose more then 100000 values it will render out results alot faster but not the actual values generated and sorted.

[toc] | [prev] | [next] | [standalone]

#34229

Den söndag 29 oktober 2017 kl. 05:27:47 UTC+1 skrev jonas.t...@gmail.com:
> My sort Algorithm is at least 5 time faster then native and quicks sort for densely populated samples "where numbers have 1 or more occurences".
> 
> 
> But usual there is some weird aspect i can't find out what i do wrong, if you set startrange for the random numbergererator higher then 0 values will turn up undefined i don't get it.
> 
> But the timings are correct, 5 times faster for 100k of integers, and best of all the algorithm is linear, but it purely for sorting numbers.
> 
> You guys probably see right away what goes wrong, when you change the start range. I do not.
> 
> http://anybase.co.nf/jtsort.html

<HTML> 
<HEAD><TITLE>SORT</TITLE></HEAD> 
<BODY> 

<TABLE  BGCOLOR="WHITE" BORDER=1><TH>GENERATE RANDOM NUMBERS</TH> <TR><TD>
<FORM NAME=random onSubmit="newrandom(); return false;"> 
Smallest number <input type="text" name="startrange" value="0"><br>
Biggest number <input type="text" name="endrange" value="99000"><br>
Values to generate <input type="text" name="xvalues" value="99000"><P>
<input type="submit" value="GENERATE"><P>
VALUES TO SORT "PRINTOUT ONLY FOR ARRAY WITH LESS THEN 100000 ELEMENTS"<BR> 
<TEXTAREA NAME=sort COLS=120 ROWS=8>
</TEXTAREA><BR> 
</FORM>
</TD></TR></TABLE>

<TABLE BGCOLOR="WHITE" BORDER=!>
<TH>COMPARE SORT TIMINGS</TH>
<TR><TD>CHOOSE SORT ALGORITHM
<select id="mySelect" onchange="chosesort();">
  <option value="index">JT_INDEXSORT</option>
  <option value="javascript">javascript_sort</option>
  <option value="quick">QUICKSORT</option>
</select>
<FORM NAME=sort onSubmit="chosesort(); return false;"> 
<input type=submit value="SORT"> TIME MS <input type="text" name="functime" value=""><br>
SORTED VALUES PRINTOUT "ONLY IF ARRAY HAS LESS THEN 100000 ELEMENTS":<BR>
<TEXTAREA TYPE=TEXT NAME=distrib COLS=120 ROWS=23>
</TEXTAREA><BR> 
</FORM> 
</TD></TR></TABLE>
</BODY>

<SCRIPT LANGUAGE="Javascript"> 
newrandom();
chosesort();

function newrandom(){
dval=new Array();
start=document.random.startrange.value;
end=document.random.endrange.value;
start=Number(start);
end=Number(end);
numberofvalues=document.random.xvalues.value;
//Generate random values
for (var i=0;i<numberofvalues;i++){
  number=Math.floor(Math.random() * end) + start; 
  dval[i]=number;
}
if (dval.length<100000) {document.random.sort.value=dval;} else {document.random.sort.value=dval.length+" elements in array"};
}

function chosesort(){

selectedsort = document.getElementById("mySelect").value;
if (selectedsort=="index"){jtindexsort();}
else if (selectedsort=="javascript"){javascriptNativesort();}
else if (selectedsort=="bubble"){bubbleSort();}
else if (selectedsort=="quick"){ quickSort(dval,0,numberofvalues);}
if (dval.length<100000) {printout();} else {document.sort.distrib.value=dval.length+" sorted in array"};
}

function  quickSort(arr, left, right)
{
var starttime = new Date().getTime();
	var i = left;
	var j = right;
	var tmp;
	pivotidx = (left + right) / 2; 
	var pivot = parseInt(arr[pivotidx.toFixed()]);  
	/* partition */
	while (i <= j)
	{
		while (parseInt(arr[i]) < pivot)
		i++;
		while (parseInt(arr[j]) > pivot)
			j--;
		if (i <= j)
		{
			tmp = arr[i];
			arr[i] = arr[j];
			arr[j] = tmp;
			i++;
			j--;
		}
	}

	/* recursion */
	if (left < j)
		quickSort(arr, left, j);
	if (i < right)
		quickSort(arr, i, right);
        var endtime = new Date().getTime();
        var timed = endtime - starttime;
        document.sort.functime.value=timed;
	return arr;
}

function javascriptNativesort(){
var starttime = new Date().getTime();
dval = dval.sort(sortNumber);
var endtime = new Date().getTime();
var timed = endtime - starttime;
document.sort.functime.value=timed;
}

function sortNumber(a, b)
{
  return a - b;
}

function jtindexsort(){ 
var starttime = new Date().getTime();
countval= new Array(); 
//Initiate array range
for (j=start;j<end;j++){ 
countval[j]=0; 
} 
var temp;

//Put them on correct index
for (var i=0;i<numberofvalues;i++){ 
  temp=dval[i]; 
  countval[temp]=countval[temp]+1; 
} 
p=0;
dval = [];

for (var j=start;j<end;j++){ 
   temp=countval[j] 
   for (var k=0;k<temp;k++){ 
    dval[p]=j; 
    p++;
   } 
}

var endtime = new Date().getTime();
var timed = endtime - starttime;
document.sort.functime.value=timed;
}

function printout(){
valdist="";
for (var j=0;j<numberofvalues;j++){
  valdist+=dval[j]+",";
}
document.sort.distrib.value=valdist; 
}
</SCRIPT>

</HTML>

[toc] | [prev] | [next] | [standalone]

#34230

jonas.thornvall@gmail.com writes:

> My sort Algorithm is at least 5 time faster then native and quicks
> sort for densely populated samples "where numbers have 1 or more
> occurences".

It's not really your algorithm, it's your implementation of a well-known
algorithm.

> But usual there is some weird aspect i can't find out what i do wrong,
> if you set startrange for the random numbergererator higher then 0
> values will turn up undefined i don't get it.

You don't zero the right number of elements in countval.  This, in turn,
is because you do

  number=Math.floor(Math.random() * end) + start;

so the numbers range from start to start+end-1.

> But the timings are correct, 5 times faster for 100k of integers, and
> best of all the algorithm is linear, but it purely for sorting
> numbers.

Quoting a factor is not very helpful.  It's likely that you are doing
some form of counting sort, so you can make it any number of times
faster simply by sorting a larger set.  Equally, you can probably make
it lower by choosing a small set with a huge range.

-- 
Ben.

[toc] | [prev] | [next] | [standalone]

#34231

Den söndag 29 oktober 2017 kl. 13:10:59 UTC+1 skrev Ben Bacarisse:
> jonas.thornvall@gmail.com writes:
> 
> > My sort Algorithm is at least 5 time faster then native and quicks
> > sort for densely populated samples "where numbers have 1 or more
> > occurences".
> 
> It's not really your algorithm, it's your implementation of a well-known
> algorithm.
> 
> > But usual there is some weird aspect i can't find out what i do wrong,
> > if you set startrange for the random numbergererator higher then 0
> > values will turn up undefined i don't get it.
> 
> You don't zero the right number of elements in countval.  This, in turn,
> is because you do
> 
>   number=Math.floor(Math.random() * end) + start;
> 
> so the numbers range from start to start+end-1.
> 
> > But the timings are correct, 5 times faster for 100k of integers, and
> > best of all the algorithm is linear, but it purely for sorting
> > numbers.
> 
> Quoting a factor is not very helpful.  It's likely that you are doing
> some form of counting sort, so you can make it any number of times
> faster simply by sorting a larger set.  Equally, you can probably make
> it lower by choosing a small set with a huge range.
> 
> -- 
> Ben.

Well it is my algorithm invented out of bucket sort that was known in the 80's. This was not known because back then the dinosaurs did think different because of memory requirment and so on. But you see average Joe did it already back then because they did not understand better and when they saw that it was really good for big samples with a small range they knew it was faster and better then quicksort. But it was never appreciated.

So it really is my algorithm "although one can't own algorithms i invented it", i was stupid enough to implement it to sort integers on a spectrum.

[toc] | [prev] | [next] | [standalone]

#34232

Den söndag 29 oktober 2017 kl. 13:53:38 UTC+1 skrev jonas.t...@gmail.com:
> Den söndag 29 oktober 2017 kl. 13:10:59 UTC+1 skrev Ben Bacarisse:
> > jonas.thornvall@gmail.com writes:
> > 
> > > My sort Algorithm is at least 5 time faster then native and quicks
> > > sort for densely populated samples "where numbers have 1 or more
> > > occurences".
> > 
> > It's not really your algorithm, it's your implementation of a well-known
> > algorithm.
> > 
> > > But usual there is some weird aspect i can't find out what i do wrong,
> > > if you set startrange for the random numbergererator higher then 0
> > > values will turn up undefined i don't get it.
> > 
> > You don't zero the right number of elements in countval.  This, in turn,
> > is because you do
> > 
> >   number=Math.floor(Math.random() * end) + start;
> > 
> > so the numbers range from start to start+end-1.
> > 
> > > But the timings are correct, 5 times faster for 100k of integers, and
> > > best of all the algorithm is linear, but it purely for sorting
> > > numbers.
> > 
> > Quoting a factor is not very helpful.  It's likely that you are doing
> > some form of counting sort, so you can make it any number of times
> > faster simply by sorting a larger set.  Equally, you can probably make
> > it lower by choosing a small set with a huge range.
> > 
> > -- 
> > Ben.
> 
> Well it is my algorithm invented out of bucket sort that was known in the 80's. This was not known because back then the dinosaurs did think different because of memory requirment and so on. But you see average Joe did it already back then because they did not understand better and when they saw that it was really good for big samples with a small range they knew it was faster and better then quicksort. But it was never appreciated.
> 
> So it really is my algorithm "although one can't own algorithms i invented it", i was stupid enough to implement it to sort integers on a spectrum.

Well it could been a pet using basic in the 70's

[toc] | [prev] | [next] | [standalone]

#34236

jonas.thornvall@gmail.com writes:

> Den söndag 29 oktober 2017 kl. 13:10:59 UTC+1 skrev Ben Bacarisse:
>> jonas.thornvall@gmail.com writes:
>> 
>> > My sort Algorithm is at least 5 time faster then native and quicks
>> > sort for densely populated samples "where numbers have 1 or more
>> > occurences".
>> 
>> It's not really your algorithm, it's your implementation of a well-known
>> algorithm.
<snip>
> Well it is my algorithm invented out of bucket sort that was known in
> the 80's. This was not known because back then the dinosaurs did think
> different because of memory requirment and so on. But you see average
> Joe did it already back then because they did not understand better
> and when they saw that it was really good for big samples with a small
> range they knew it was faster and better then quicksort. But it was
> never appreciated.

I don't see how you can claim to know the literature well enough to make
grandiose claims like this.  Knuth (in TAOCP) says it was invented in
the early 1950s.  I knew it well in 1970s though I could not have told
you then how invented it.

I think you should speak to someone you trust about these thoughts.
It's not normal for someone starting out in a field to believe they have
invented significant things hitherto not spotted by "the dinosaurs" but
which, nonetheless, go unappreciated.

<snip>
-- 
Ben.

[toc] | [prev] | [next] | [standalone]

#34238

Den söndag 29 oktober 2017 kl. 15:47:14 UTC+1 skrev Ben Bacarisse:
> jonas.thornvall@gmail.com writes:
> 
> > Den söndag 29 oktober 2017 kl. 13:10:59 UTC+1 skrev Ben Bacarisse:
> >> jonas.thornvall@gmail.com writes:
> >> 
> >> > My sort Algorithm is at least 5 time faster then native and quicks
> >> > sort for densely populated samples "where numbers have 1 or more
> >> > occurences".
> >> 
> >> It's not really your algorithm, it's your implementation of a well-known
> >> algorithm.
> <snip>
> > Well it is my algorithm invented out of bucket sort that was known in
> > the 80's. This was not known because back then the dinosaurs did think
> > different because of memory requirment and so on. But you see average
> > Joe did it already back then because they did not understand better
> > and when they saw that it was really good for big samples with a small
> > range they knew it was faster and better then quicksort. But it was
> > never appreciated.
> 
> I don't see how you can claim to know the literature well enough to make
> grandiose claims like this.  Knuth (in TAOCP) says it was invented in
> the early 1950s.  I knew it well in 1970s though I could not have told
> you then how invented it.
> 
> I think you should speak to someone you trust about these thoughts.
> It's not normal for someone starting out in a field to believe they have
> invented significant things hitherto not spotted by "the dinosaurs" but
> which, nonetheless, go unappreciated.
> 
> <snip>
> -- 
> Ben.

You are silly Ben the 1950's computers used transistor and radio tubes, the memories was mercury delay line programs and data was either on card or switches.

Well you can ask the guys that renovated Edsac to implement my algorithm, they have to be very fast flipping the switches read the data in LoL

Well your a funny guy Ben, not that bright but funny ;)

[toc] | [prev] | [next] | [standalone]

#34241

jonas.thornvall@gmail.com writes:

> Den söndag 29 oktober 2017 kl. 15:47:14 UTC+1 skrev Ben Bacarisse:
>> jonas.thornvall@gmail.com writes:
>> 
>> > Den söndag 29 oktober 2017 kl. 13:10:59 UTC+1 skrev Ben Bacarisse:
>> >> jonas.thornvall@gmail.com writes:
>> >> 
>> >> > My sort Algorithm is at least 5 time faster then native and quicks
>> >> > sort for densely populated samples "where numbers have 1 or more
>> >> > occurences".
>> >> 
>> >> It's not really your algorithm, it's your implementation of a well-known
>> >> algorithm.
>> <snip>
>> > Well it is my algorithm invented out of bucket sort that was known in
>> > the 80's. This was not known because back then the dinosaurs did think
>> > different because of memory requirment and so on. But you see average
>> > Joe did it already back then because they did not understand better
>> > and when they saw that it was really good for big samples with a small
>> > range they knew it was faster and better then quicksort. But it was
>> > never appreciated.
>> 
>> I don't see how you can claim to know the literature well enough to make
>> grandiose claims like this.  Knuth (in TAOCP) says it was invented in
>> the early 1950s.  I knew it well in 1970s though I could not have told
>> you then how invented it.
<snip>
> You are silly Ben the 1950's computers used transistor and radio
> tubes, the memories was mercury delay line programs and data was
> either on card or switches.

No, it's silly of you to question a very well-researched work like
Knuth's TAOCP.  Harold Seward's in 1954 PhD thesis is even available
online.  You can see the algorithm there as pain as day.  You think your
limited knowledge of computer history is all there is.  Some algorithms
were even invented before there were digital computers.

> Well you can ask the guys that renovated Edsac to implement my
> algorithm, they have to be very fast flipping the switches read the
> data in LoL

You don't need a computer to have an algorithm, but as it happens,
Harold Seward's work was motivated by having very little fast memory.
You therefore need a method that works well with very slow memory.  A
counting sort is very good for that situation.

> Well your a funny guy Ben, not that bright but funny ;)

Have you heard of the Dunning-Kruger effect?

-- 
Ben.

[toc] | [prev] | [next] | [standalone]

#34242

Den söndag 29 oktober 2017 kl. 16:46:47 UTC+1 skrev Ben Bacarisse:
> jonas.thornvall@gmail.com writes:
> 
> > Den söndag 29 oktober 2017 kl. 15:47:14 UTC+1 skrev Ben Bacarisse:
> >> jonas.thornvall@gmail.com writes:
> >> 
> >> > Den söndag 29 oktober 2017 kl. 13:10:59 UTC+1 skrev Ben Bacarisse:
> >> >> jonas.thornvall@gmail.com writes:
> >> >> 
> >> >> > My sort Algorithm is at least 5 time faster then native and quicks
> >> >> > sort for densely populated samples "where numbers have 1 or more
> >> >> > occurences".
> >> >> 
> >> >> It's not really your algorithm, it's your implementation of a well-known
> >> >> algorithm.
> >> <snip>
> >> > Well it is my algorithm invented out of bucket sort that was known in
> >> > the 80's. This was not known because back then the dinosaurs did think
> >> > different because of memory requirment and so on. But you see average
> >> > Joe did it already back then because they did not understand better
> >> > and when they saw that it was really good for big samples with a small
> >> > range they knew it was faster and better then quicksort. But it was
> >> > never appreciated.
> >> 
> >> I don't see how you can claim to know the literature well enough to make
> >> grandiose claims like this.  Knuth (in TAOCP) says it was invented in
> >> the early 1950s.  I knew it well in 1970s though I could not have told
> >> you then how invented it.
> <snip>
> > You are silly Ben the 1950's computers used transistor and radio
> > tubes, the memories was mercury delay line programs and data was
> > either on card or switches.
> 
> No, it's silly of you to question a very well-researched work like
> Knuth's TAOCP.  Harold Seward's in 1954 PhD thesis is even available
> online.  You can see the algorithm there as pain as day.  You think your
> limited knowledge of computer history is all there is.  Some algorithms
> were even invented before there were digital computers.
> 
> > Well you can ask the guys that renovated Edsac to implement my
> > algorithm, they have to be very fast flipping the switches read the
> > data in LoL
> 
> You don't need a computer to have an algorithm, but as it happens,
> Harold Seward's work was motivated by having very little fast memory.
> You therefore need a method that works well with very slow memory.  A
> counting sort is very good for that situation.
> 
> > Well your a funny guy Ben, not that bright but funny ;)
> 
> Have you heard of the Dunning-Kruger effect?
> 
> -- 
> Ben.

Nope should i?

[toc] | [prev] | [next] | [standalone]

#34245

In article <c14e005f-4f60-4b9c-9427-eabef1ff495f@googlegroups.com>,
<jonas.thornvall@gmail.com> wrote:

>Den söndag 29 oktober 2017 kl. 16:46:47 UTC+1 skrev Ben Bacarisse:

>> Have you heard of the Dunning-Kruger effect?

>Nope should i?

Well it applies to you jonas:

  <https://en.wikipedia.org/wiki/Dunning–Kruger_effect>

-- 
27/6/1975 - Herbert Kiebler shot and killed trying to cross Berlin Wall.

"A reminder that the defining characteristic of a socialist regime is coercion,
not equality."

Dan Hannan

[toc] | [prev] | [next] | [standalone]

#34243

Den söndag 29 oktober 2017 kl. 16:46:47 UTC+1 skrev Ben Bacarisse:
> jonas.thornvall@gmail.com writes:
> 
> > Den söndag 29 oktober 2017 kl. 15:47:14 UTC+1 skrev Ben Bacarisse:
> >> jonas.thornvall@gmail.com writes:
> >> 
> >> > Den söndag 29 oktober 2017 kl. 13:10:59 UTC+1 skrev Ben Bacarisse:
> >> >> jonas.thornvall@gmail.com writes:
> >> >> 
> >> >> > My sort Algorithm is at least 5 time faster then native and quicks
> >> >> > sort for densely populated samples "where numbers have 1 or more
> >> >> > occurences".
> >> >> 
> >> >> It's not really your algorithm, it's your implementation of a well-known
> >> >> algorithm.
> >> <snip>
> >> > Well it is my algorithm invented out of bucket sort that was known in
> >> > the 80's. This was not known because back then the dinosaurs did think
> >> > different because of memory requirment and so on. But you see average
> >> > Joe did it already back then because they did not understand better
> >> > and when they saw that it was really good for big samples with a small
> >> > range they knew it was faster and better then quicksort. But it was
> >> > never appreciated.
> >> 
> >> I don't see how you can claim to know the literature well enough to make
> >> grandiose claims like this.  Knuth (in TAOCP) says it was invented in
> >> the early 1950s.  I knew it well in 1970s though I could not have told
> >> you then how invented it.
> <snip>
> > You are silly Ben the 1950's computers used transistor and radio
> > tubes, the memories was mercury delay line programs and data was
> > either on card or switches.
> 
> No, it's silly of you to question a very well-researched work like
> Knuth's TAOCP.  Harold Seward's in 1954 PhD thesis is even available
> online.  You can see the algorithm there as pain as day.  You think your
> limited knowledge of computer history is all there is.  Some algorithms
> were even invented before there were digital computers.
> 
> > Well you can ask the guys that renovated Edsac to implement my
> > algorithm, they have to be very fast flipping the switches read the
> > data in LoL
> 
> You don't need a computer to have an algorithm, but as it happens,
> Harold Seward's work was motivated by having very little fast memory.
> You therefore need a method that works well with very slow memory.  A
> counting sort is very good for that situation.
> 
> > Well your a funny guy Ben, not that bright but funny ;)
> 
> Have you heard of the Dunning-Kruger effect?
> 
> -- 
> Ben.

You see Ben one do not invent algorithms that superior and never use them and teach them. 

That is how i know it "BULL" so to speak, well you get what i mean....

[toc] | [prev] | [next] | [standalone]

#34244

Den söndag 29 oktober 2017 kl. 16:46:47 UTC+1 skrev Ben Bacarisse:
> jonas.thornvall@gmail.com writes:
> 
> > Den söndag 29 oktober 2017 kl. 15:47:14 UTC+1 skrev Ben Bacarisse:
> >> jonas.thornvall@gmail.com writes:
> >> 
> >> > Den söndag 29 oktober 2017 kl. 13:10:59 UTC+1 skrev Ben Bacarisse:
> >> >> jonas.thornvall@gmail.com writes:
> >> >> 
> >> >> > My sort Algorithm is at least 5 time faster then native and quicks
> >> >> > sort for densely populated samples "where numbers have 1 or more
> >> >> > occurences".
> >> >> 
> >> >> It's not really your algorithm, it's your implementation of a well-known
> >> >> algorithm.
> >> <snip>
> >> > Well it is my algorithm invented out of bucket sort that was known in
> >> > the 80's. This was not known because back then the dinosaurs did think
> >> > different because of memory requirment and so on. But you see average
> >> > Joe did it already back then because they did not understand better
> >> > and when they saw that it was really good for big samples with a small
> >> > range they knew it was faster and better then quicksort. But it was
> >> > never appreciated.
> >> 
> >> I don't see how you can claim to know the literature well enough to make
> >> grandiose claims like this.  Knuth (in TAOCP) says it was invented in
> >> the early 1950s.  I knew it well in 1970s though I could not have told
> >> you then how invented it.
> <snip>
> > You are silly Ben the 1950's computers used transistor and radio
> > tubes, the memories was mercury delay line programs and data was
> > either on card or switches.
> 
> No, it's silly of you to question a very well-researched work like
> Knuth's TAOCP.  Harold Seward's in 1954 PhD thesis is even available
> online.  You can see the algorithm there as pain as day.  You think your
> limited knowledge of computer history is all there is.  Some algorithms
> were even invented before there were digital computers.
> 
> > Well you can ask the guys that renovated Edsac to implement my
> > algorithm, they have to be very fast flipping the switches read the
> > data in LoL
> 
> You don't need a computer to have an algorithm, but as it happens,
> Harold Seward's work was motivated by having very little fast memory.
> You therefore need a method that works well with very slow memory.  A
> counting sort is very good for that situation.
> 
> > Well your a funny guy Ben, not that bright but funny ;)
> 
> Have you heard of the Dunning-Kruger effect?
> 
> -- 
> Ben.

You see this is so easy a 10 year old could implement it, and it is way faster but it was never teached to anyone so noone ever used it to sort LoL

The general problem for you ENTP guys doing problem solving is you lack the ability to think outside the box. You can only build your knowledge on already existing approaches that you master. You are the caretakers om information theoretical approaches...... Not the invontors......
Yeah you make sense Ben, "BULL" sense

[toc] | [prev] | [next] | [standalone]

#34246

jonas.thornvall@gmail.com writes:
<snip>
> You see this is so easy a 10 year old could implement it, and it is
> way faster but it was never teached to anyone so noone ever used it to
> sort LoL

What?  Now you claim to know what is taught.  It is in TAOPC and it was
taught in the university I taught at.  Unlike you, I can't claim to be
an expert in what other universities teach.

> The general problem for you ENTP guys doing problem solving is you
> lack the ability to think outside the box. You can only build your
> knowledge on already existing approaches that you master. You are the
> caretakers om information theoretical approaches...... Not the
> invontors......  Yeah you make sense Ben, "BULL" sense

And you now claim to know if I am or am not creative.  Is there really
no limit to what you think you know?

-- 
Ben.

[toc] | [prev] | [next] | [standalone]

#34247

Den söndag 29 oktober 2017 kl. 17:18:35 UTC+1 skrev Ben Bacarisse:
> jonas.thornvall@gmail.com writes:
> <snip>
> > You see this is so easy a 10 year old could implement it, and it is
> > way faster but it was never teached to anyone so noone ever used it to
> > sort LoL
> 
> What?  Now you claim to know what is taught.  It is in TAOPC and it was
> taught in the university I taught at.  Unlike you, I can't claim to be
> an expert in what other universities teach.
> 
> > The general problem for you ENTP guys doing problem solving is you
> > lack the ability to think outside the box. You can only build your
> > knowledge on already existing approaches that you master. You are the
> > caretakers om information theoretical approaches...... Not the
> > invontors......  Yeah you make sense Ben, "BULL" sense
> 
> And you now claim to know if I am or am not creative.  Is there really
> no limit to what you think you know?
> 
> -- 
> Ben.

Well Ben i actually thought you knew about Myer Briggs, maybe you should check it out. Well i have no fear for my intellects i know my shortcomings i am sloppy but intelligent far above Mensa qualities. More like on in a million.

[toc] | [prev] | [next] | [standalone]

#34248

Den söndag 29 oktober 2017 kl. 17:18:35 UTC+1 skrev Ben Bacarisse:
> jonas.thornvall@gmail.com writes:
> <snip>
> > You see this is so easy a 10 year old could implement it, and it is
> > way faster but it was never teached to anyone so noone ever used it to
> > sort LoL
> 
> What?  Now you claim to know what is taught.  It is in TAOPC and it was
> taught in the university I taught at.  Unlike you, I can't claim to be
> an expert in what other universities teach.
> 
> > The general problem for you ENTP guys doing problem solving is you
> > lack the ability to think outside the box. You can only build your
> > knowledge on already existing approaches that you master. You are the
> > caretakers om information theoretical approaches...... Not the
> > invontors......  Yeah you make sense Ben, "BULL" sense
> 
> And you now claim to know if I am or am not creative.  Is there really
> no limit to what you think you know?
> 
> -- 
> Ben.

The thing is you do not make sense Ben, what you say do not make sense, that is why i call it bull Ben. If there is a faster algorithm that is at least a multiple of times faster then Quicksort and so easy to learn that you could teach it to a 10 year old. Why isn't it used, why was it not used i have done my share of record sorting using quicksort you should know. 

So i know it wasn't teached and noone i saw used it at university.

[toc] | [prev] | [next] | [standalone]

#34250

jonas.thornvall@gmail.com writes:

> Den söndag 29 oktober 2017 kl. 17:18:35 UTC+1 skrev Ben Bacarisse:
>> jonas.thornvall@gmail.com writes:
>> <snip>
>> > You see this is so easy a 10 year old could implement it, and it is
>> > way faster but it was never teached to anyone so noone ever used it to
>> > sort LoL
>> 
>> What?  Now you claim to know what is taught.  It is in TAOPC and it was
>> taught in the university I taught at.  Unlike you, I can't claim to be
>> an expert in what other universities teach.
<snip>
> So i know it wasn't teached and noone i saw used it at university.

How do you know that?

-- 
Ben.

[toc] | [prev] | [next] | [standalone]

#34253

Den söndag 29 oktober 2017 kl. 19:02:08 UTC+1 skrev Ben Bacarisse:
> jonas.thornvall@gmail.com writes:
> 
> > Den söndag 29 oktober 2017 kl. 17:18:35 UTC+1 skrev Ben Bacarisse:
> >> jonas.thornvall@gmail.com writes:
> >> <snip>
> >> > You see this is so easy a 10 year old could implement it, and it is
> >> > way faster but it was never teached to anyone so noone ever used it to
> >> > sort LoL
> >> 
> >> What?  Now you claim to know what is taught.  It is in TAOPC and it was
> >> taught in the university I taught at.  Unlike you, I can't claim to be
> >> an expert in what other universities teach.
> <snip>
> > So i know it wasn't teached and noone i saw used it at university.
> 
> How do you know that?
> 
> -- 
> Ben.

Well we had one or two algorithm classes, when it come to speed there was nothing else then quicksort and maybe bucketsort in rare cases.

[toc] | [prev] | [next] | [standalone]

Page 1 of 2  [1] 2  Next page →

Back to top | Article view | comp.lang.javascript

csiph-web