CRLF line terminators -> LF line terminators.
catseye
7 years ago

0 | The Jaccia and Jacciata Cellular Automata | |

1 | ========================================= | |

2 | ||

3 | Overview | |

4 | -------- | |

5 | ||

6 | Jaccia and Jacciata are cellular automata inspired by the Announcement of | |

7 | [Scientific Proof that Slime Molds are Intelligent Maze Solvers][]. | |

8 | ||

9 | Basically, the slime mold solves the maze by: | |

10 | ||

11 | - initially being everywhere in the maze | |

12 | - there being food at the entrance and exit of the maze | |

13 | - minimizing its surface area by retreating from anywhere it can't get | |

14 | food. | |

15 | ||

16 | Jaccia operates the same way. In the Jaccia automaton, slime cells | |

17 | survive if they have immediate neighbours in at least two cardinal | |

18 | directions that provide sustenance, i.e. are either food or other slime | |

19 | cells. The result is the same: paths of slime cells that lead down dead | |

20 | ends have one end which provides no sustenance and dies off. Eventually, | |

21 | only paths of slime cells that reach from food to food (or uninterrupted | |

22 | circular paths of slime cells) remain. Jacciata is a more involved | |

23 | automaton which finds only the shortest path. | |

24 | ||

25 | [Scientific Proof that Slime Molds are Intelligent Maze Solvers]: http://web.archive.org/web/20020220163303/http://www.riken.go.jp/lab-www/frontier-div/NEWSLETTER/feb2001/ameboid_e.htm | |

26 | ||

27 | Properties | |

28 | ---------- | |

29 | ||

30 | Jaccia has the property that, when started from this condition (entire | |

31 | maze filled with slime cells), the automaton will eventually reach a | |

32 | fixed point (steady state) which contains all possible orthogonal paths | |

33 | from food to food. (Orthogonal paths means, a diagonal isn't considered | |

34 | a path.) | |

35 | ||

36 | Jacciata is similar, but has the property that when it reaches a fixed | |

37 | point, it will contain the *shortest* path from food to food, if such a | |

38 | path exists and is unique. If no such path exists, or is not unique, the | |

39 | result is undefined. It is otherwise similar to Jaccia. | |

40 | ||

41 | Definition | |

42 | ---------- | |

43 | ||

44 | Both Jaccia and Jacciata are defined in ALPACA v1.0. Jaccia is defined | |

45 | in the file `jaccia.alp` and Jacciata in `jacciata.alp`. The ALPACA | |

46 | definition is authoritative; what is given here is merely advisory. | |

47 | ||

48 | Both automata use basically the same set of symbols. An initial Jaccia | |

49 | playfield generally serves as an initial Jacciata playfield with the | |

50 | same kind of solution. | |

51 | ||

52 | - ` ` - empty space | |

53 | - `#` - wall (purely decorative) | |

54 | - `%` - slime mold | |

55 | - `F` - food | |

56 | - `S` - "start" food (needed in Jacciata, optional in Jaccia) | |

57 | - `-` - exploratory head (Jacciata only) | |

58 | - `?` - exploratory body (Jacciata only) | |

59 | - `@` - solved (Jacciata only) | |

60 | ||

61 | Discussion | |

62 | ---------- | |

63 | ||

64 | Jacciata's definition is not very elegant, especially when compared to | |

65 | Jaccia. In order for it to work, the two sources of food need to be | |

66 | labelled differently (`S` and `F`), there needs to be a "head" of an | |

67 | exploratory shoot that looks for solutions, and so on. It could probably | |

68 | be made more elegant with some work. | |

69 | ||

70 | [New in 1.1] The definition of these automata in ALPACA 0.94 suggested some | |

71 | possible improvements to ALPACA, particularly the definition of | |

72 | neighbourhoods different from the assumed von Neumann neighbourhood, and | |

73 | their use in the count operator. The Jaccia and Jacciata descriptions were | |

74 | rewritten in ALPACA 1.0, and do now take advantage of these features in order | |

75 | to be written more succinctly. | |

76 | ||

77 | Happy intelligence! Such as it is. | |

78 | Chris Pressey | |

79 | April 11, 2009 | |

80 | Bellevue, WA | |

0 | The Jaccia and Jacciata Cellular Automata | |

1 | ========================================= | |

2 | ||

3 | Overview | |

4 | -------- | |

5 | ||

6 | Jaccia and Jacciata are cellular automata inspired by the Announcement of | |

7 | [Scientific Proof that Slime Molds are Intelligent Maze Solvers][]. | |

8 | ||

9 | Basically, the slime mold solves the maze by: | |

10 | ||

11 | - initially being everywhere in the maze | |

12 | - there being food at the entrance and exit of the maze | |

13 | - minimizing its surface area by retreating from anywhere it can't get | |

14 | food. | |

15 | ||

16 | Jaccia operates the same way. In the Jaccia automaton, slime cells | |

17 | survive if they have immediate neighbours in at least two cardinal | |

18 | directions that provide sustenance, i.e. are either food or other slime | |

19 | cells. The result is the same: paths of slime cells that lead down dead | |

20 | ends have one end which provides no sustenance and dies off. Eventually, | |

21 | only paths of slime cells that reach from food to food (or uninterrupted | |

22 | circular paths of slime cells) remain. Jacciata is a more involved | |

23 | automaton which finds only the shortest path. | |

24 | ||

25 | [Scientific Proof that Slime Molds are Intelligent Maze Solvers]: http://web.archive.org/web/20020220163303/http://www.riken.go.jp/lab-www/frontier-div/NEWSLETTER/feb2001/ameboid_e.htm | |

26 | ||

27 | Properties | |

28 | ---------- | |

29 | ||

30 | Jaccia has the property that, when started from this condition (entire | |

31 | maze filled with slime cells), the automaton will eventually reach a | |

32 | fixed point (steady state) which contains all possible orthogonal paths | |

33 | from food to food. (Orthogonal paths means, a diagonal isn't considered | |

34 | a path.) | |

35 | ||

36 | Jacciata is similar, but has the property that when it reaches a fixed | |

37 | point, it will contain the *shortest* path from food to food, if such a | |

38 | path exists and is unique. If no such path exists, or is not unique, the | |

39 | result is undefined. It is otherwise similar to Jaccia. | |

40 | ||

41 | Definition | |

42 | ---------- | |

43 | ||

44 | Both Jaccia and Jacciata are defined in ALPACA v1.0. Jaccia is defined | |

45 | in the file `jaccia.alp` and Jacciata in `jacciata.alp`. The ALPACA | |

46 | definition is authoritative; what is given here is merely advisory. | |

47 | ||

48 | Both automata use basically the same set of symbols. An initial Jaccia | |

49 | playfield generally serves as an initial Jacciata playfield with the | |

50 | same kind of solution. | |

51 | ||

52 | - ` ` - empty space | |

53 | - `#` - wall (purely decorative) | |

54 | - `%` - slime mold | |

55 | - `F` - food | |

56 | - `S` - "start" food (needed in Jacciata, optional in Jaccia) | |

57 | - `-` - exploratory head (Jacciata only) | |

58 | - `?` - exploratory body (Jacciata only) | |

59 | - `@` - solved (Jacciata only) | |

60 | ||

61 | Discussion | |

62 | ---------- | |

63 | ||

64 | Jacciata's definition is not very elegant, especially when compared to | |

65 | Jaccia. In order for it to work, the two sources of food need to be | |

66 | labelled differently (`S` and `F`), there needs to be a "head" of an | |

67 | exploratory shoot that looks for solutions, and so on. It could probably | |

68 | be made more elegant with some work. | |

69 | ||

70 | [New in 1.1] The definition of these automata in ALPACA 0.94 suggested some | |

71 | possible improvements to ALPACA, particularly the definition of | |

72 | neighbourhoods different from the assumed von Neumann neighbourhood, and | |

73 | their use in the count operator. The Jaccia and Jacciata descriptions were | |

74 | rewritten in ALPACA 1.0, and do now take advantage of these features in order | |

75 | to be written more succinctly. | |

76 | ||

77 | Happy intelligence! Such as it is. | |

78 | Chris Pressey | |

79 | April 11, 2009 | |

80 | Bellevue, WA |