Rather than proposing a fancy sorting algorithm, I'd suggest reviewing the
Quine-McKlusky algotithm. If every boolean equation (or those with about
12 inputs) required sorting 8 million entries then it would take forever to
compile hdl code. I wrote a Q-M routine in Pascal about 15 years ago and I don't
recall having to do this. It was so damned long ago that the only thing I can
remember about it were that I nearly went insane implement my first recursive
program.

recursion - n. See the definition of recursion.

When you generate your mappings of your terms, try looking for a clever way of
ordering the data. It may save you a little time...

Scott

On Wed, 23 Aug 2000, Samuel Winchenbach wrote:

> >Could you better explain what you are trying to do.  Would the number
> >00001111 come before 00011110, or when there is the same number of ones
> >is there a secondary sort to bring the binary implementation into
> >play?
>
> *SNIP*
>
> Thanks for the response.
>
> 0000
> 0001
> 0010
> 0011
> 0100
> 0101
> 0110
> 0111
> 1000
> 1001
> 1010
> 1011
> 1100
> 1101
> 1110
> 1111
>
> would be sorted to:
>
> 0000
>
> 0001
> 0010
> 0100
> 1000
>
> 0011
> 0101
> 0110
> 1001
> 1010
> 1100
>
> 0111
> 1011
> 1101
> 1110
>
> 1111
>
>
> As you can see the numbers are sorted by the number of one's in the binary
> representation.  The actual value has nothing to do with it.  So what I
> have done is the following:
>
>
> class quine
> {
>   private:
>
>   public:
>     int value;
>     int blanks;
>
>     bool operator > (quine &rightside) const;
>     bool operator >= (quine &rightside) const;
>     bool operator < (quine &rightside) const;
>     bool operator <= (quine &rightside) const;
>     bool operator == (quine &rightside) const;
> };
>
> value is the actual number (NOT the number of ones in the binary
> representaion)
>
> next I store all 8 million numbers in a vector of quine:
>
> vector<quine> mterms;  //minterms
>
> for( blah blah.. )
> {
>   //cycle throught the text file, and push a type "quine" into the vector
>   //after setting its value.
> }
>
> now the next thing I want to do is sort my vector of type quine.  not by
> value, but by the number of one's in the binary representation of value.
>
> I already have this working using selection sort.... but by god, I think I
> would grow old and die before it finished.  Radix or Quicksort seem the
> best bet... but I have no idea how to implment them... and nothing from
> the STL will help me either (I guess sort in <algorithm> is quicksort..
> but I do not believe it will work due to my strange data stucture.
>
> Thanks again,
>
> Sam
>
> --
> http://www.piclist.com hint: The list server can filter out subtopics
> (like ads or off topics) for you. See http://www.piclist.com/#topics
>
>

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu