网络标识模型的扩散与社会效用函数

Question

我正在写一个关于思想在加权有向图上传播的网络标志模型。主要的概念是传播红色的想法:代理的重量超过0.5将变成红色。我还需要绘制一个实用程序函数，我正在考虑这样做：

plot sum [p] of turtles with [color = red] * sum [weight] of links with [color = red and weight > 0.5] / (sum [p] of turtles * sum [weight] of links with [weight > 0.5])

例如，每个代理的所有实用价值之和(p * w)。然而，我认为这是相当静态的，我需要一些建议。我想用一种更好的方法来增加p的值，增加每只海龟从一只红海龟那里得到的加权值，或者通过减去每只海龟从一只灰海龟那里得到的加权值来降低p的值。

有人能帮我实现这一点吗？

extensions [nw] ;network extention
breed [group g] ;turtles are called group
directed-link-breed [influences inf] ;directed link are called influences
links-own [weight] ;weight of directed link, related to links only
group-own [p] ;probability to take or not red idea, related to turtles only

SETUP

;generation of the "real world" networks

to setup-pa ;generate preferential-attachment network
  ca
  nw:generate-preferential-attachment group influences members 1 [
    set color grey
    setxy random-xcor random-ycor
    set shape "person"
    set size 2]
  assign-weight
  assign-p
  reset-ticks
end

to setup-small-world ;generate small-world network
  ca
  nw:generate-watts-strogatz group influences members 1 1 [
    set color grey
    setxy random-xcor random-ycor
    set shape "person"
    set size 2]
  assign-weight
  assign-p
  reset-ticks
end

to setup-random ;generate random network
  ca
  nw:generate-random group influences members 1 [
    set color grey
    setxy random-xcor random-ycor
    set shape "person"
    set size 2]
  assign-weight
  assign-p
  reset-ticks
end

to assign-weight ;assign random-float weight to links in a range from 0 to 1
  ask influences [set weight random-float 1 ]
end

to assign-p ;assign probability to agents (0 or 1)
  ask group [set p random 2]
end


EXPERIMENT

to behavior
  ask turtles with [p = 0] [ ;asking members with p = 0 and with out-links' weight > 0.5 to spread grey colour
    set color grey
    ask my-out-links with [weight > 0.5] [set color grey ask end2 [set color grey]]]
    ask turtles with [p = 1] [ ;asking members with p=1 and with out-links' weight > 0.5 to spread red idea
    set color red
    ask my-out-links with [weight > 0.5] [set color red ask end2 [set color red]]]
  tick
end

enter code here

Answer 1

Excuse the sloppy code, I'm sick. I changed the behavior block

对符号输入之和，ala神经网络。对于100个成员和随机种子12345，稳定需要5个步骤。许多调试行，很抱歉它是如此草率，但它运行。

韦德

extensions [nw] ;network extention
breed [group g] ;turtles are called group
directed-link-breed [influences inf] ;directed link are called influences
links-own [weight] ;weight of directed link, related to links only
group-own [p oldsum] ;probability to take or not red idea, related to turtles only

;;SETUP
globals [ changing   ]

;; stopping conditons
to setcounts
  set changing true
  set members 100  ;; if not in a slider
end


to setseed
  ;; for members = 100 and seed = 12345 we get diffusion for 5 ticks
  random-seed 12345
end

to setup-pa ;generate preferential-attachment network
  ca
  setseed
  setcounts
  nw:generate-preferential-attachment group influences members 1 [
    set color grey
    setxy random-xcor random-ycor
    set shape "person"
    set size 2]
  assign-weight
  assign-p
  reset-ticks
end

to setup-small-world ;generate small-world network
  ca
  setseed
  setcounts
  nw:generate-watts-strogatz group influences members 1 1 [
    set color grey
    setxy random-xcor random-ycor
    set shape "person"
    set size 2]
  assign-weight
  assign-p
  reset-ticks
end

to setup-random ;generate random network
  ca
  setseed
  setcounts
  nw:generate-random group influences members 1 [
    set color grey
    setxy random-xcor random-ycor
    set shape "person"
    set size 2]
  assign-weight
  assign-p
  reset-ticks
end

to assign-weight ;assign random-float weight to links in a range from 0 to 1
  ask influences [set weight ( round (1000 * random-float 1) / 1000) ]   ;; avoid tiny variations
end

to assign-p ;assign probability to agents (0 or 1)
  ;;;;;;;; default color grey
  ask group [set p random 2  ]

  ;;;;;;; added next lines.  Change p to +1 or -1
  ask group [ if (p = 1) [ set color red ]  ]
  ask group [ if (p = 0) [ set p -1 set color grey]]
end


;;EXPERIMENT
;; change influence going OUTWARD to influence coming INWARD
;; change single link weights as determinants to SUM of link weights as determinants, as in a neural net

to go
  if (changing = false) [ print "no longer changing, so stopping." stop]
  if (ticks > 20) [stop]

  set changing  false
  print (word "=================== tick # " ticks " =====================================")
  ask links [ set label round (100 * weight) set thickness (weight) ]
  ask links [ if [color = red ] of end1 [set color red ]]
  ask turtles   [
     ;;print (word "\n ------ " who "-----")
     let summer 0
     ask my-in-links [


         ;;;;; let netweight ( weight * [ p value of end1 ]
         let netweight ( weight * ( [ p ] of end1 )  )
         set summer ( summer + netweight)
         ;;show summer
    ]
     set oldsum precision oldsum 3
     set summer precision summer 3

    if  (oldsum != summer) [   set changing true ]
       if  (oldsum != summer) and (ticks > 2)  [ print (word who "..CHANGED SUM, was " oldsum ", now " summer )
        set changing true ]
    set oldsum summer
    let oldcolor color
    (ifelse
         (summer > 0.3)  [ set p  1  set color red
                            if ( color != oldcolor) [ print (word who " turned red") ]
                         ]
         (summer < -0.3) [ set p -1  set color grey
                            if ( color != oldcolor) [ print (word who " turned grey") ]
                          ]
         ;; otherwise, do nothing
    )

  ]



  tick
end

;;enter code here