| 1 | >,[
|
|---|
| 2 | [
|
|---|
| 3 | ----------[
|
|---|
| 4 | >>>[>>>>]+[[-]+<[->>>>++>>>>+[>>>>]++[->+<<<<<]]<<<]
|
|---|
| 5 | ++++++[>------<-]>--[>>[->>>>]+>+[<<<<]>-],<
|
|---|
| 6 | ]>
|
|---|
| 7 | ]>>>++>+>>[
|
|---|
| 8 | <<[>>>>[-]+++++++++<[>-<-]+++++++++>[-[<->-]+[<<<<]]<[>+<-]>]
|
|---|
| 9 | >[>[>>>>]+[[-]<[+[->>>>]>+<]>[<+>[<<<<]]+<<<<]>>>[->>>>]+>+[<<<<]]
|
|---|
| 10 | >[[>+>>[<<<<+>>>>-]>]<<<<[-]>[-<<<<]]>>>>>>>
|
|---|
| 11 | ]>>+[[-]++++++>>>>]<<<<[[<++++++++>-]<.[-]<[-]<[-]<]<,
|
|---|
| 12 | ]
|
|---|
| 13 |
|
|---|
| 14 | [The Collatz problem or 3n+1 problem is as follows. Take a natural number n.
|
|---|
| 15 | If it's even, halve it; if odd, triple it and add one. Repeat the process with
|
|---|
| 16 | the resulting number, and continue indefinitely. If n is 0, the resulting
|
|---|
| 17 | sequence is 0, 0, 0, 0... It is conjectured but not proven that for any
|
|---|
| 18 | positive integer n, the resulting sequence will end in 1, 4, 2, 1...
|
|---|
| 19 | See also http://www.research.att.com/projects/OEIS?Anum=A006577
|
|---|
| 20 |
|
|---|
| 21 | This program takes a series of decimal numbers, followed by linefeeds (10).
|
|---|
| 22 | The entire series is terminated by an EOF (0 or "no change"). For each number
|
|---|
| 23 | input, the program outputs, in decimal, the number of steps from that number
|
|---|
| 24 | to zero or one, when following the rule above. It's quite fast; on a Sun
|
|---|
| 25 | machine, it took three seconds for a random 640-digit number.
|
|---|
| 26 |
|
|---|
| 27 | One more note. This program was originally written for Tristan Parker's
|
|---|
| 28 | Brainfuck Texas Holdem contest, and won by default (it was the only entry);
|
|---|
| 29 | the version I submitted before the contest deadline is at
|
|---|
| 30 | http://www.hevanet.com/cristofd/brainfuck/oldcollatz.b
|
|---|
| 31 |
|
|---|
| 32 | Daniel B Cristofani (cristofdathevanetdotcom)
|
|---|
| 33 | http://www.hevanet.com/cristofd/brainfuck/]
|
|---|
