Neuron Simulation

Classes

Node

• coor: (x, y) coordinates.
• connections: number of neighbours, can be 2, 3, 4, 6, or 8.
• neighbours: return the neighbour nodes of current node. Number of neighbours is determined by self.connections.
• __hash__(): operator overloading, used for comparison and in set.
• __eq__(): operator overloading, used for comparison.
• __repr__(): operator overloading, representation.

Path

• origin: one of the nodes the path is connecting. Origin node.
• dest: the other node the path is connecting. Destination node. Grow direction of path is self.origin -> self.dest.
• length: length of path.
• alive: if True, the path can grow at next step, otherwise it cannot grow.
• max_length: the max allowed length of current path. The path will be divided into 2 or more paths in Neuron.clean() if self.length > self.max_length.
• died(): Turn a path to dead state by setting self.alive to False.
• grow(): increase self.length.
• __hash__(): operator overloading, used for comparison and in set.
• __eq__(): operator overloading, used for comparison.
• __repr__(): operator overloading, representation.

Neuron

A neuron is a collection of Nodes and Paths. It grows as time passes.

Basic properties

• origin: the neuron's starting node.
• speed: the grow speed for each path.
• vertex: number of neighbours for each node.
• hand: number of neurites one neuron has at the beginning of a simulation.
• split_prob: probability for one neurite to split into 2 neurites on a node. A neurite will split only when it encounters a new node.

Conformation properties

• connected: True if the neuron is connected to any other neurons.
• nodes: nodes that are occupied by one neuron.
• boundary_nodes: nodes that are in the grow direction of one neuron.
• paths: paths that are fully occupied by one neuron.
• boundary_paths: paths that are partially occupied by one neuron.

Methods

• born(): initiate bare neuron with hands. Update boundary_paths and boundary_nodes.
• grow(): increase the length of one neuron's alive boundary_paths by amount of speed.
• clean(): check if any path in boundary_paths has exceeded its maximum length. If one path exceeds its maximum length:
  1. include the destination node to current neuron's nodes.
  2. Decide whether to split or not by using split_check().
  3. Decide which way to grow by using way_to_go().
  4. Include new boundary paths and boundary nodes to boundary_paths and boundary_nodes.
• connect(): Turn self.connected to True.
• check_alive(): Check paths in boundary_paths if they are alive or not.
• split_check(): Decide whether to split at a node by using self.split_prob. Return True or False to decide whether to split and possible boundary nodes for way_to_go() to choose.
• way_to_go(): Decide which way to go when encountering a new node. Just choose from possible destination nodes split_check() passes.
• cal_end(self, path): calculate coordinates of a endpoint for a path.
• draw(): Draw origin node and current paths on a Draw object.

Exp

Simulation class placeholder.

Functions

check_connection()

1. Check if any two neurons are connected. If two neurons are connected, call their connect() method to update their status.
2. Connections can only occur between boundary nodes.
3. Checking method:
   • Store a list of connected tuples (neuron1, neuron2), do not check for these combinations.
   • For each combination n1 and n2:
     1. Check n1.nodes and n2.nodes for duplication. If there are duplications, mark n1 and n2 as connected and stop checking. If no duplications, go to step 2.
     2. Check for duplications between n1.boundary_nodes and n2.boundary_nodes. If there are duplications, say n1.boundary_node and n2.boundary_node, let path1 be the boundary path for n1 that ends with n2.boundary_node, and path2 be the boundary path for n2 that ends with n1.boundary_node. Check the sum for path1.length and path2.length. If the sum exceeds maximum path length, mark n1 and n2 as connected, else stop checking.

stats_connections()

Count the percentage of connected neurons.

coor_equal()

Check equality for two pairs of float coordinates.

Patterns

pattern42()

Generate pattern grid for type 2 and type 4. Then randomly select nodes as neuron origin nodes.

pattern6()

Generate pattern grid for type 6. Then randomly select nodes as neuron origin nodes.

Process

1. Initiate grid and neurons.
2. T = 0, For each neuron, born().
3. T > 0, for each time step:
   1. For each neuron grow().
   2. For each neuron clean().
   3. For each neuron check_alive().
   4. check_connections().
4. If percentage of connected neurons exceeds the pre-defined max-percentage, stop iteration.