DIM a (Max Index)

[SMcNeill]

What I'm really curious about, at this point, is how do we know that there's always going to be less than 256 occurrences of a number?  You're limiting the counter to a single unsigned byte, but it seems like billions of numbers generated would have a chance of at least one of them occurring more than 255 times.

[SMcNeill]

The best solution:

Code: QB64: [Select]

1. $CHECKING:OFF
2. DIM q AS _INTEGER64
3. DIM TimesProcessed AS _UNSIGNED _BYTE
4.  
5. limit = 1000000
6. DIM m AS _MEM
7. m = _MEMNEW(9000000000)
8. _MEMFILL m, m.OFFSET, 9000000000, 0 AS _UNSIGNED _BYTE
9.  
10. RANDOMIZE TIMER
11. starttime## = TIMER
12. FOR i = 1 TO limit
13.     FOR j = 1 TO 31 'for 31 q's
14.         q = INT(RND * 9000000000)
15.         _MEMGET m, m.OFFSET + q, TimesProcessed
16.         TimesProcessed = TimesProcessed + 1
17.         _MEMPUT m, m.OFFSET + q, TimesProcessed
18.     NEXT
19. NEXT
20. finishtime## = TIMER - starttime##
21. PRINT USING "#####.### seconds for ###,###,###,### numbers"; finishtime##, 31 * limit

A few cavets here:

1) You need a 64-bit version of QB64 to use this.
2) You need 9GB of ram or so.

On my machine, we process 31,000,000 values in 5 seconds, so I'd guess a billion numbers would take less than 5 minutes. (Compared with 40 seconds for 3,100 values using file storage.)

[TempodiBasic]

@Steve
you've used the direct access to the memory! I think that I cannot get so fast results.
I'm no expert to use RAM in this low level mode!
Great lesson.

[soniaa]

I have Upgraded the RAM in my Pc;
(old; at the start of this post: 6 GB)
(old; at the middle of this post: 8 GB)
NOW: 12 GB

[SMcNeill]

I have Upgraded the RAM in my Pc;
(old; at the start of this post: 6 GB)
(old; at the middle of this post: 8 GB)
NOW: 12 GB

Try a method like the one in reply #46.  Use _MEMNEW and _MEMPUT/_MEMGET to write the values directly to memory.

[soniaa]

Thank you so much,,,  to all of you !!!

What seemed to me a good way (open c.dat for binary);
I found out to be not good,
when the records stored in this file they become 1 billion, GET e PUT it slows down a lot,
...probably because this method does not use much RAM.

[soniaa]

- - - - My Problem Is Solved. - - - - -

I've just finished a test;

I inserted in my program the method:      REDIM rec(9000000000) as _UNSIGNED _BYTE
The results  they were excellent !!!  .....much more expected !!!
The speed reached is enormous, my program has runned in phases where I had never-never-never been able to arrive.
The estimated time for to get to the end   it was  2 to 6  years first,

--- NOW The estimated time for to get to the end  it is  1 or 2 hours !!! ---