I can't decide which I like best...

[bplus]

From reply #2

Code: QB64: [Select]

1. WIDTH 144, 25
2. _SCREENMOVE 0, 0
3. j = 0
4. DO
5.     READ mydata$
6.     IF mydata$ = "EOF" THEN EXIT DO
7.     j = j + 1
8. LOOP
9. RESTORE
10. DIM myarray$(j)
11.  
12. DO
13.     READ mydata$
14.     IF mydata$ = "EOF" THEN EXIT DO
15.     j = 1: occ% = 0: h = h + 1
16.     dataconcat$ = mydata$ + "|" + dataconcat$
17.     DO UNTIL j > LEN(dataconcat$)
18.         IF INSTR(j, dataconcat$, mydata$ + "|") - j = 0 THEN occ% = occ% + 1
19.         j = INSTR(j, dataconcat$, "|") + 1
20.     LOOP
21.     myarray$(h) = mydata$ + "#" + LTRIM$(STR$(occ%))
22. LOOP
23. dataconcat$ = ""
24. FOR i = 1 TO h
25.     FOR j = 1 TO h
26.         IF i <> j THEN
27.             IF myarray$(i) < myarray$(j) THEN SWAP myarray$(i), myarray$(j)
28.         END IF
29. NEXT j, i
30. FOR i = 1 TO h
31.     PRINT myarray$(i); ", ";
32. NEXT
33. PRINT "End Method #1."
34.  
35. DATA dog,cat,rabbit,frog,horse,dog,mouse,pig,cat,bat,cat,dog,bird,fish,cat,pig,dog
36. DATA EOF
37.  

Trimmed reply #2 down a bit:

Code: QB64: [Select]

1. DIM myarray$(1 TO 1000) ' some over the top number to avoid counting and then DIM
2. DO
3.     READ mydata$
4.     IF mydata$ = "EOF" THEN EXIT DO
5.     nItems = nItems + 1: myarray$(nItems) = mydata$
6. LOOP
7. FOR i = 1 TO nItems - 1 'nice sort routine, does this always work?
8.     FOR j = i + 1 TO nItems
9.         IF myarray$(i) > myarray$(j) THEN SWAP myarray$(i), myarray$(j)
10. NEXT j, i
11. FOR i = 1 TO nItems ' if we really must number the animals
12.     IF i = 1 THEN
13.         lastWord$ = myarray$(i): cnt = 1
14.     ELSE
15.         IF myarray$(i) = lastWord$ THEN cnt = cnt + 1 ELSE cnt = 1: lastWord$ = myarray$(i)
16.     END IF
17.     myarray$(i) = lastWord$ + "#" + _TRIM$(STR$(cnt))
18.     IF i = nItems THEN PRINT myarray$(i) ELSE PRINT myarray$(i); ", ";
19. NEXT
20. DATA dog,cat,rabbit,frog,horse,dog,mouse,pig,cat,bat,cat,dog,bird,fish,cat,pig,dog,EOF
21.  

I modified your sort routine, man that is short and sweet! Does it always work?

[bplus]

Looks cute:

Code: QB64: [Select]

1. DIM b(1 TO 480) 'lilSort test b+ 2020-09-23
2. DO
3.     nItems = INT(RND * (480)) + 1
4.     FOR i = 1 TO nItems
5.         b(i) = INT(RND * 100)
6.     NEXT
7.     lilSort b(), nItems
8.     FOR i = 1 TO nItems
9.         PRINT b(i);
10.     NEXT
11.     SLEEP
12.     CLS
13. LOOP UNTIL _KEYDOWN(27)
14.  
15. SUB lilSort (arr(), topItem)
16.     DIM i AS INTEGER, j AS INTEGER
17.     FOR i = 1 TO topItem - 1
18.         FOR j = i + 1 TO topItem
19.             IF arr(i) > arr(j) THEN SWAP arr(i), arr(j)
20.     NEXT j, i
21. END SUB
22.  
23.  

In action, I suppose it is about the worst in efficiency:

Code: QB64: [Select]

1. DIM b(1 TO 480) 'lilSort test b+ 2020-09-23
2. DO
3.     nItems = INT(RND * (480)) + 1
4.     FOR i = 1 TO nItems
5.         b(i) = INT(RND * 100)
6.     NEXT
7.     lilSort b(), nItems
8.     CLS
9.     FOR i = 1 TO nItems
10.         PRINT b(i);
11.     NEXT
12.     _DISPLAY
13.     SLEEP
14.     CLS
15. LOOP UNTIL _KEYDOWN(27)
16.  
17. SUB lilSort (arr(), topItem)
18.     DIM i AS INTEGER, j AS INTEGER
19.     FOR i = 1 TO topItem - 1
20.         FOR j = i + 1 TO topItem
21.             IF arr(i) > arr(j) THEN SWAP arr(i), arr(j)
22.             GOSUB show
23.             _LIMIT 800
24.     NEXT j, i
25.     EXIT SUB
26.     show:
27.     CLS
28.     FOR k = 1 TO topItem
29.         PRINT arr(k);
30.     NEXT
31.     _DISPLAY
32.     RETURN
33. END SUB
34.  
35.  

[SMcNeill]

Nice, Mark.

My way to add an array and place the results in alphabetical order would be as follows...

Code: QB64: [Select]

1. WIDTH 144, 25
2. _SCREENMOVE 0, 0
3. j = 0
4. DO
5.     READ mydata$
6.     IF mydata$ = "EOF" THEN EXIT DO
7.     j = j + 1
8. LOOP
9. RESTORE
10. DIM myarray$(j)
11.  
12. DO
13.     READ mydata$
14.     IF mydata$ = "EOF" THEN EXIT DO
15.     j = 1: occ% = 0: h = h + 1
16.     dataconcat$ = mydata$ + "|" + dataconcat$
17.     DO UNTIL j > LEN(dataconcat$)
18.         IF INSTR(j, dataconcat$, mydata$ + "|") - j = 0 THEN occ% = occ% + 1
19.         j = INSTR(j, dataconcat$, "|") + 1
20.     LOOP
21.     myarray$(h) = mydata$ + "#" + LTRIM$(STR$(occ%))
22. LOOP
23. dataconcat$ = ""
24. FOR i = 1 TO h
25.     FOR j = 1 TO h
26.         IF i <> j THEN
27.             IF myarray$(i) < myarray$(j) THEN SWAP myarray$(i), myarray$(j)
28.         END IF
29. NEXT j, i
30. FOR i = 1 TO h
31.     PRINT myarray$(i); ", ";
32. NEXT
33. PRINT "End Method #1."
34.  
35. DATA dog,cat,rabbit,frog,horse,dog,mouse,pig,cat,bat,cat,dog,bird,fish,cat,pig,dog
36. DATA EOF
37.  

Pete

Bubblesort??    AAAAARRRRRGGGHHHHHH!!!  My poor eyes!  My poor programmer’s sensibilities!!  Bubble sort is even worse coding practice than GOTO!!!

And the ABSOLUTE worst version of Bubblesort!  You didn’t even make the inner loop FOR j = i + 1 TO h...

I...  I feel faint!  Sorry guys...  I gotta go barf.  Somebody please teach Pete a decent sort algorithm!  CODEGUY, WE NEED YOU, QUICK!!!

Blaaaarrrrrggggghhhhhh...

[SpriggsySpriggs]

Here is my contribution which works quite nicely. When timed on my machine for sorting and rebuilding the array it took 0.31 seconds.

Code: QB64: [Select]

1. _TITLE "Sort da list"
2. DO
3.     READ mydata$
4.     IF mydata$ = "EOF" THEN EXIT DO
5.     nItems = nItems + 1
6.     REDIM _PRESERVE myarray$(nItems)
7.     myarray$(nItems) = mydata$
8. LOOP
9.  
10. OPEN "datatosort.txt" FOR OUTPUT AS #1
11. FOR x = 1 TO UBOUND(myarray$)
12.     PRINT #1, myarray$(x)
13. NEXT
14. CLOSE #1
15.  
16. SHELL _HIDE "PowerShell Get-Content datatosort.txt | sort | Out-File sorteddata.txt -Encoding ASCII" 'regular sort
17. 'SHELL _HIDE "PowerShell Get-Content datatosort.txt | sort | Get-Unique | Out-File sorteddata.txt -Encoding ASCII" 'sort with duplicates omitted
18.  
19. OPEN "sorteddata.txt" FOR BINARY AS #1
20. FOR x = 1 TO UBOUND(myarray$)
21.     IF NOT EOF(1) THEN
22.         LINE INPUT #1, myarray$(x)
23.         PRINT myarray$(x)
24.     END IF
25. NEXT
26. CLOSE #1
27.  
28. DATA dog,cat,rabbit,frog,horse,dog,mouse,pig,cat,bat,cat,dog,bird,fish,cat,pig,dog,EOF

  [ You are not allowed to view this attachment ]    [ You are not allowed to view this attachment ]  

Code: QB64: [Select]

1. _TITLE "Sort da list"
2. x# = TIMER(0.01)
3. SHELL _HIDE "PowerShell Get-Content 9999words.txt | sort | Out-File sorteddata.txt -Encoding ASCII"
4. 'SHELL _HIDE "PowerShell Get-Content 9999words.txt | sort | Get-Unique | Out-File sorteddata.txt -Encoding ASCII"
5.  
6. OPEN "sorteddata.txt" FOR BINARY AS #1
7. DO
8.     IF EOF(1) = 0 THEN
9.         x = x + 1
10.         REDIM _PRESERVE myarray$(x)
11.         LINE INPUT #1, myarray$(x)
12.     END IF
13. LOOP UNTIL EOF(1)
14. CLOSE #1
15. y# = TIMER(0.01)
16. PRINT y# - x#

The above code along with my attachment "9999words.txt" can be sorted in 0.77 seconds on my machine and if you want only unique you can get it in 0.56 seconds.

[Pete]

Bad salsa on Taco Tuesday again, Steve? I hope you recover in time for Pon Farr Friday, but in case you don't, check out my fruit sort at QBF sometime. Ah, the good ol' days. Anyway, the speed test isn't for the alphabetizing part, it is for determining which of the two methods of using MID$() or INSTR() work faster for determining duplicates during the concatenation process.

Oh dammit, I think I stepped in some...

Pete

[STxAxTIC]

I lolled at Steve's reaction to bubble sort. I felt myself being channeled from somewhere so I had to check the forums for flames, and sure enough... Haha

[bplus]

Come on guys, as Pete says you don't need a ballistic missile to kill a deer specially if you want to eat it after.

Remember this is for Pete's phone calls, you don't need the heavy artillery for a couple dozen items to sort.

.31 secs!!! Ha!

Check this out

Code: QB64: [Select]

1. start = TIMER(.001)
2. WIDTH 144, 25
3. '_SCREENMOVE 0, 0  '<<< try the time commented and not commented out
4. j = 0
5. DO
6.     READ mydata$
7.     IF mydata$ = "EOF" THEN EXIT DO
8.     j = j + 1
9. LOOP
10. RESTORE
11. DIM myarray$(j)
12.  
13. DO
14.     READ mydata$
15.     IF mydata$ = "EOF" THEN EXIT DO
16.     j = 1: occ% = 0: h = h + 1
17.     dataconcat$ = mydata$ + "|" + dataconcat$
18.     DO UNTIL j > LEN(dataconcat$)
19.         IF INSTR(j, dataconcat$, mydata$ + "|") - j = 0 THEN occ% = occ% + 1
20.         j = INSTR(j, dataconcat$, "|") + 1
21.     LOOP
22.     myarray$(h) = mydata$ + "#" + LTRIM$(STR$(occ%))
23. LOOP
24. dataconcat$ = ""
25. FOR i = 1 TO h
26.     FOR j = 1 TO h
27.         IF i <> j THEN
28.             IF myarray$(i) < myarray$(j) THEN SWAP myarray$(i), myarray$(j)
29.         END IF
30. NEXT j, i
31. FOR i = 1 TO h
32.     PRINT myarray$(i); ", ";
33. NEXT
34. PRINT "End Method #1."
35. PRINT "Time:"; TIMER(.001) - start
36. DATA dog,cat,rabbit,frog,horse,dog,mouse,pig,cat,bat,cat,dog,bird,fish,cat,pig,dog
37. DATA EOF
38.  
39.  
40.  

Discovery: the line that takes the most time by far in this program is the _SCREENMOVE !!!

I don't think Pete even cares about the sorting, I brought in the subject because I thought it easier to # the animals after a sort.

BTW I have a version of fruit sort, cool!

Code: QB64: [Select]

1. RANDOMIZE TIMER
2. DEFINT A-Z
3. numberOfFruits = 25
4. DIM Basket$(10)
5. FOR i = 1 TO numberOfFruits 'create fruits
6.     Fruit = INT(RND * 9) + 1: f$ = RIGHT$(STR$(Fruit), LEN(STR$(Fruit)) - 1)
7.     IF i <> numberOfFruits THEN PRINT f$; ", "; ELSE PRINT f$;
8.     Basket$(Fruit) = Basket$(Fruit) + f$ + ", "
9. NEXT
10. PRINT: PRINT
11. FOR Nut = 1 TO 9
12.     IF LEN(Basket$(Nut)) THEN
13.         IF Nut <> 9 THEN PRINT Basket$(Nut); ELSE PRINT LEFT$(Basket$(Nut), LEN(Basket$(Nut)) - 2)
14.     END IF
15. NEXT
16.  
17.  

[SpriggsySpriggs]

.31 secs!!! Ha!

@bplus That is because on smaller data sets, PowerShell will be slower. However, if you look at my updated reply, I can sort 9999 words in 0.77 seconds or 0.56 seconds if I omit duplicates. Try using my list of 9999 words and see how your algorithm does.

[bplus]

@bplus That is because on smaller data sets, PowerShell will be slower. However, if you look at my updated reply, I can sort 9999 words in 0.77 seconds or 0.56 seconds if I omit duplicates. Try using my list of 9999 words and see how your algorithm does.

Sure 9999 words but then I would pull out Qsort and blow that time away and then you will say, "Wait let's try 999,999,999 words then Powershell will win! :-))

I am saying for a couple hundred words even the worst sort is better than jobbing out to Powereshell or whatever Hydrogen Bomb of the sorting world you want to bring in.

[Pete]

That's why I stick with my patented Stupid Sort. It's the quickest, easiest sort code to remember and type into the IDE. I can place it in a program in under 30 seconds. It works plenty fast for most jobs where the lists to sort are 1000 items or less, but you guys probably thought I posted it just to make Steve puke. Nah, that's just a bonus!

Pete

 - If you're out of your mind, see a psychiatrist. If you're just out of sorts, see CodeGuy.

[SMcNeill]

That's why I stick with my patented Stupid Sort. It's the quickest, easiest sort code to remember and type into the IDE. I can place it in a program in under 30 seconds. It works plenty fast for most jobs where the lists to sort are 1000 items or less, but you guys probably thought I posted it just to make Steve puke. Nah, that's just a bonus!

Pete

 - If you're out of your mind, see a psychiatrist. If you're just out of sorts, see CodeGuy.

That’s why I keep several sorts in my folder, and all I need do is either $INCLUDE them, or copy/paste them into my program.  It’s efficient program-wise, and even faster coding than typing a puke-sort is.  ;D