view release on metacpan or  search on metacpan

README  view on Meta::CPAN

        *~/.config/cellgraph* in your home directory. It contains mainly
        stored colors and dirs where to load and store setting files. You
        may change it manually or deleted it to reset it to default.

DESCRIPTION
    This graphical application uses cellular automata logic, as described in
    *Steve Wolfram*s book *"A new kind of science"*, to paint tiled
    pictures. Although, the original concept got expanded by many additional
    options and functionalities.

    It is meant for fun, leasure, beautiful, personalized images and a
    deeper understanding about how cellular automatons work.

Mechanics
    Every tile (square) in the picture represents one automaton (called
    cell). The tile color depicts the state of that cell. The state is just
    an one digit number (0 or 1 at start). To see and change which color
    stands for which state - choose the rightmost tab titled "*Colors*". The
    uppermost row of the picture represents a row of automata in its initial
    state, that is given by the user via settings in the second tab
    ("*Starting Row*"). The row below is painted by the same string of
    cells, just after one round of computation later and so forth. The
    vertical Y-axis is so to speak the time axis, with top being the
    beginning and bottom the end.

    During each round of computation every cell might change its state. It
    depends on which subrule matches. Each subrule is layed out as a row at
    the third tab ("*State Rules*"). On a left side of the arrow (=>) you
    see there a number of colored tiles. At the beginning its three tiles,
    representing our focal automaton in the middle and its left neighbour on
    its left flank and its right neighbour on the right. In case all three
    colors around our example cell are as displayed, then the new state of
    this cell can be read on the right side of the arrow. Its also called
    the result of this sub-rule. Of course this was simplified, since many
    options might complicate that picture, as decribed in next paragraphs.

    One big addition are action rules. Parallel to its state, every cell has
    also an activity value. It starts with a value also set in the
    "*Starting Row*" tab. It drops every round by a fixed amount but it also
    gets raised depending on the result of action rules. Only if the
    activity value reaches a threshold, the cell can change its state. More
    details about this mechanics are to be found in the chapter "*General
    Settings*" and "*Action Rules*".

GUI
    The general layout is very simple: the picture gets drawn on the left
    window side. On the right side you change the settings from which the
    picture is computed. Please note there the tabs, in the top row. They
    select which page of settings is visible.

    Please mind the tool tips - short help texts which appear if the mouse
    stands still over a button. Also helpful are messages in the status bar
    at the bottom that appear while browsing the menu of after a command
    given.

  General Settings
    The first tab contains settings, that shape the drawing in the most
    broad way. It is segmented into three parts that somewhat parallel the
    last three tabs.

    The topmost section sets the framework for rules by which the cell state
    changes - computation round by computation round. Input Size appoints
    the size of neighbourhood, the left side of an subrule. If you set it to
    an odd number like 5, then the cells current state plus its two
    neighbours on each side determine the next state of a cell. But if you
    set it to an even number like 2 then only one neighbour on each side has
    this power, but not the focal cell itself. You can recognize this by the
    struck through middle cell in each subrule in the "*State Rules*" tab.
    Cell States sets the number of different cell states / colors. The
    Select option defines an subrule mapping - meaning: how many of the
    possible subrule results you can define manually? This aimes to solve
    the none trivial problem of combinatorial explosion. To guide you in
    that decision which mapping to choose - there are also two read only
    text fields in the second row, that display how many rules and subrules
    are possible during current settings. Let's clarify with an example: If
    you got 5 *Cell State*s and an *Input Size* of 3, each of these 3
    neighbourhood cells can be in one of 5 states. Since 5 to the power 3 is
    125 , thats how many subrules get displayed in tab thee and four. For a
    lot of people this is no longer convenient. Other combinations like 8 **
    5 = 32768 would just overload the programm (caution when set these
    values!). But given the example of 5 ** 3 = 125. 125 subrules can be
    managed. But since every subrule can have any state as result - there
    are 5 ** 125 = (way too many) subrule combinations = rules possible. The
    rule number display in tab 1 and 3 stop working but the rest of app
    still churns on. To reduce the subrule count, you could set the subrule
    mapping from *all* to *symmetric*. Symmetric twins like the subrule
    input pattern 123 and 321 would have the then the same result. Only the
    pattern 123 (left neighbour has state 1, cell has 2 and right neighbour
    3) will be displayed - 75 subrules in total. If that is still too much,
    select subrule mapping *sorted*. Then also 132, 213, 231 and 312 belong
    to the same group, because if sorted they all are equal to 123. Now we
    got only 35 subrules. The most compression you get with the mapping
    named *summing*. Since 1+2+3 = 6, all input pattern that result in the
    same sum belong to the same group and there will be only 13 groups left.
    And furthermore 5**13 = 1_220_703_125 - all displays can work as
    intended. And you stil have 1_220_703_125 pattern to choose from. Please
    note if you select any mapping other than *all* the picture becomes
    mirror-symmetric. The option Result is normally set to *insert*, meaning
    the new state defined by the matching subrule will be inserted = is the
    new cell state. But for sake of variation you could also add, subtract
    or multiply the new state with the previous state. The outcome of that
    operation modulo cell state count will become the new state. Two options
    contain the suffix _rot which means an additional one wil be added so
    you rotate through the states even if all subrules are not set and
    blank. Cells on the left and right edge normally have only a reduced
    amount of neighbours. During computation the virtual left neighbour of
    the left most cell has always the state zero. But if you activate the
    option Circular, then the left neighbour of the leftmost cell is the
    rightmost cell and vice versa. This can fix certain types of
    irregularities in the drawing.

    The middle section sets the framework for the action rules, which change
    the activity value of a cell. The activity value can never be below zero
    or above one. The Apply option activates the aplication of action rules.
    And if action rules are in effect then the state of a cell can only
    change, if the activity value is equal or above the Threshold. Action
    rules react to the same input pattern as state rules. However, the
    result of an action subrule is the increase of the activity value by an
    amount that is different for any subrule that can have its own result
    state. This amount of increase is usually but doesn't have to be
    positive. There is also another amount which is usually negative and
    which decreases the activity value of all cells each round. This
    decrease value is labeled here Change and has to be negative in order to
    decrease all activity values. Then there is also a Spread value, which
    is set to zero per default. If set to two, the a cell can influence the
    activity value of the neext two neigbours on the left and right. How
    much will also determined in the action rule tab. Since ever ection rule
    has also associated a second value. This second value defines how much a
    cell can influence its outer most neighbours (which value gets added to
    them). The neighbours in between the out most neighbour and the pivotal
    cell get influence by an amount that is linearly interpolated.

    The bottom section is about visual settings, which sometimes are not
    just cosmetic. The Direction helps you to draw completely different
    pictures with way more symmetry. If set to *top_down* (default) the
    picture gets painted as described in the "Mechanics" paragraph (above).
    But if you set it to *outside_in*, then you will see only a triangular
    slice of the previous pattern, painted four times. Every outer edge of
    the square shaped map will be the first row, displaying the same pattern
    and the computation will grow toward the center. The option *inside_out*
    is a kinda opposite. Here it starts in the center of the grid growing
    outward in all four directions. If you deactivate Fill, each tile is no
    longer filled by the state color. Instead only two small lines get
    drawn. Grid Style offers three options: *lines*, *gaps* and *no*. The
    first (default) option draws thin grey lines between the cells. These
    lines are white if *gaps* is chosen. And there will be *no* gaps between
    the tiles if that option is selected . And Cell Size simply defines how
    many pixel a tile edge is long.

  Starting Row
    This tab contains settings that define the start values - the states and
    activity values of all cells in the starting row. The upper part is
    about the cell states and the lower part about the activity values.
    Since both parts parallel each other, lets explain them in one go.

    Central in the upper and lower part are 20 cells that change their state
    by left or right clicking on them. You either cycle that way through the
    cell states or increase and decrease the activity value by steps of 0.2.

    Above the cell row you can read the summary of the selected values.
    Trailing zero's will get removed and to fit the value into the summary,
    each activity value will be multiplied by 5, so that 5 stands for an
    activity value of 1.0 (the maximum) of this cell and a 0 for 0 activity.
    The selected values will be placed into the center of the starting row
    in the drawing. Unless - you click on the Repeat option. Then the chosen
    patter gets repeated until the first row is full. Right beside the
    summary display are a few buttons that trigger changes in the starting
    row values. The 1 always (also in other panels) resets the default
    values. And ? always triggers a random value generator on all cells for
    happy accidents. In addition to that there are also the buttons with
    arrows on it. They just the summary value up or down, in order to give
    you the next or previous start configuration.