Взяти за основу модель матричної гри з ненульовою сумою
globals[
A
B
]
patches-own [
strategy ;; current strategy of patch
old-strategy ;; old strategy of patch
score ;; score resulting from interaction of neighboring patches
]
to setup
clear-all
ask patches [
ifelse random-float 1.0 < (initial-strategy-A / 100)
[ setup-strategy 1]
[setup-strategy 2]
establish-color
]
reset-ticks
end
to setup-strategy [value]
set strategy value
set old-strategy value
end
to go
ask patches [interact] ;; to play with a neighboring patch
ask patches [select-strategy] ;; adopt the strategy of the neighbor (who had the highest score)
tick
end
to go-once
ask patches [interact] ;; to play with a neighboring patch
ask patches [select-strategy] ;; adopt the strategy of the neighbor (who had the highest score)
tick
end
to interact ;; patch procedure
let randpatch one-of patches
let strategy1 [strategy] of randpatch
let strategy2 [strategy] of self
if (not interact?) [set randpatch one-of neighbors]
ifelse (strategy2 = 1)
[ifelse (strategy1 = 1)
[set score A-A * coordination-award]
[set score A-B]
]
[ifelse (strategy1 = 1)
[set score B-A ]
[set score B-B * coordination-award]
]
end
to select-strategy ;; patch procedure
set old-strategy strategy
set strategy [strategy] of max-one-of neighbors [score]
establish-color
end
to establish-color ;; patch procedure
ifelse (old-strategy = 1)
[ifelse (strategy = 1)
[set pcolor blue]
[set pcolor orange ]
]
[ifelse (strategy = 1)
[set pcolor grey]
[set pcolor red]
]
end
Завдання: розв’язати свій варіант
Перетворити гру на біматричну (тобто додати інший тип гравця з іншими платежами для стратегій А та B): задати платежі обох гравців, що приймають участь у взаємодії. Запропонувати власне рішення як уникнути подвійного обходу гравців.
Додати затримку на зміну стратегії, створити можливість двох популяцій з різними затримками. Дослідити зміну динаміки для різних значень затримки.
Додати третій варіант стратегій - С, змінити матрицю гри та модель відповідним чином.
Додати окрему популяцію, яка завжди використовує максимінну стратегію (для ігор з нульовою сумою).
Взяти за основу модель взаємодії у грі ділема ув’язненого (назва у бібліотеці моделей PD N person iterated)
globals [
;;number of turtles with each strategy
num-random
num-cooperate
num-defect
num-tit-for-tat
num-unforgiving
num-unknown
;;number of interactions by each strategy
num-random-games
num-cooperate-games
num-defect-games
num-tit-for-tat-games
num-unforgiving-games
num-unknown-games
;;total score of all turtles playing each strategy
random-score
cooperate-score
defect-score
tit-for-tat-score
unforgiving-score
unknown-score
]
turtles-own [
score
strategy
defect-now?
partner-defected? ;;action of the partner
partnered? ;;am I partnered?
partner ;;WHO of my partner (nobody if not partnered)
partner-history ;;a list containing information about past interactions
;;with other turtles (indexed by WHO values)
]
;;;;;;;;;;;;;;;;;;;;;;
;;;Setup Procedures;;;
;;;;;;;;;;;;;;;;;;;;;;
to setup
clear-all
store-initial-turtle-counts ;;record the number of turtles created for each strategy
setup-turtles ;;setup the turtles and distribute them randomly
reset-ticks
end
;;record the number of turtles created for each strategy
;;The number of turtles of each strategy is used when calculating average payoffs.
;;Slider values might change over time, so we need to record their settings.
;;Counting the turtles would also work, but slows the model.
to store-initial-turtle-counts
set num-random n-random
set num-cooperate n-cooperate
set num-defect n-defect
set num-tit-for-tat n-tit-for-tat
set num-unforgiving n-unforgiving
set num-unknown n-unknown
end
;;setup the turtles and distribute them randomly
to setup-turtles
make-turtles ;;create the appropriate number of turtles playing each strategy
setup-common-variables ;;sets the variables that all turtles share
end
;;create the appropriate number of turtles playing each strategy
to make-turtles
create-turtles num-random [ set strategy "random" set color gray - 1 ]
create-turtles num-cooperate [ set strategy "cooperate" set color red ]
create-turtles num-defect [ set strategy "defect" set color blue ]
create-turtles num-tit-for-tat [ set strategy "tit-for-tat" set color lime ]
create-turtles num-unforgiving [ set strategy "unforgiving" set color turquoise - 1 ]
create-turtles num-unknown [set strategy "unknown" set color magenta ]
end
;;set the variables that all turtles share
to setup-common-variables
ask turtles [
set score 0
set partnered? false
set partner nobody
setxy random-xcor random-ycor
]
setup-history-lists ;;initialize PARTNER-HISTORY list in all turtles
end
;;initialize PARTNER-HISTORY list in all turtles
to setup-history-lists
let num-turtles count turtles
let default-history [] ;;initialize the DEFAULT-HISTORY variable to be a list
;;create a list with NUM-TURTLE elements for storing partner histories
repeat num-turtles [ set default-history (fput false default-history) ]
;;give each turtle a copy of this list for tracking partner histories
ask turtles [ set partner-history default-history ]
end
;;;;;;;;;;;;;;;;;;;;;;;;
;;;Runtime Procedures;;;
;;;;;;;;;;;;;;;;;;;;;;;;
to go
clear-last-round
ask turtles [ partner-up ] ;;have turtles try to find a partner
let partnered-turtles turtles with [ partnered? ]
ask partnered-turtles [ select-action ] ;;all partnered turtles select action
ask partnered-turtles [ play-a-round ]
do-scoring
tick
end
to clear-last-round
let partnered-turtles turtles with [ partnered? ]
ask partnered-turtles [ release-partners ]
end
;;release partner and turn around to leave
to release-partners
set partnered? false
set partner nobody
rt 180
set label ""
end
;;have turtles try to find a partner
;;Since other turtles that have already executed partner-up may have
;;caused the turtle executing partner-up to be partnered,
;;a check is needed to make sure the calling turtle isn't partnered.
to partner-up ;;turtle procedure
if (not partnered?) [ ;;make sure still not partnered
rt (random-float 90 - random-float 90) fd 1 ;;move around randomly
set partner one-of (turtles-at -1 0) with [ not partnered? ]
if partner != nobody [ ;;if successful grabbing a partner, partner up
set partnered? true
set heading 270 ;;face partner
ask partner [
set partnered? true
set partner myself
set heading 90
]
]
]
end
;;choose an action based upon the strategy being played
to select-action ;;turtle procedure
if strategy = "random" [ act-randomly ]
if strategy = "cooperate" [ cooperate ]
if strategy = "defect" [ defect ]
if strategy = "tit-for-tat" [ tit-for-tat ]
if strategy = "unforgiving" [ unforgiving ]
if strategy = "unknown" [ unknown ]
end
to play-a-round ;;turtle procedure
get-payoff ;;calculate the payoff for this round
update-history ;;store the results for next time
end
;;calculate the payoff for this round and
;;display a label with that payoff.
to get-payoff
set partner-defected? [defect-now?] of partner
ifelse partner-defected? [
ifelse defect-now? [
set score (score + 1) set label 1
] [
set score (score + 0) set label 0
]
] [
ifelse defect-now? [
set score (score + 5) set label 5
] [
set score (score + 3) set label 3
]
]
end
;;update PARTNER-HISTORY based upon the strategy being played
to update-history
if strategy = "random" [ act-randomly-history-update ]
if strategy = "cooperate" [ cooperate-history-update ]
if strategy = "defect" [ defect-history-update ]
if strategy = "tit-for-tat" [ tit-for-tat-history-update ]
if strategy = "unforgiving" [ unforgiving-history-update ]
if strategy = "unknown" [ unknown-history-update ]
end
;;;;;;;;;;;;;;;;
;;;Strategies;;;
;;;;;;;;;;;;;;;;
;;All the strategies are described in the Info tab.
to act-randomly
set num-random-games num-random-games + 1
ifelse (random-float 1.0 < 0.5) [
set defect-now? false
] [
set defect-now? true
]
end
to act-randomly-history-update
;;uses no history- this is just for similarity with the other strategies
end
to cooperate
set num-cooperate-games num-cooperate-games + 1
set defect-now? false
end
to cooperate-history-update
;;uses no history- this is just for similarity with the other strategies
end
to defect
set num-defect-games num-defect-games + 1
set defect-now? true
end
to defect-history-update
;;uses no history- this is just for similarity with the other strategies
end
to tit-for-tat
set num-tit-for-tat-games num-tit-for-tat-games + 1
set partner-defected? item ([who] of partner) partner-history
ifelse (partner-defected?) [
set defect-now? true
] [
set defect-now? false
]
end
to tit-for-tat-history-update
set partner-history
(replace-item ([who] of partner) partner-history partner-defected?)
end
to unforgiving
set num-unforgiving-games num-unforgiving-games + 1
set partner-defected? item ([who] of partner) partner-history
ifelse (partner-defected?)
[set defect-now? true]
[set defect-now? false]
end
to unforgiving-history-update
if partner-defected? [
set partner-history
(replace-item ([who] of partner) partner-history partner-defected?)
]
end
;;defaults to tit-for-tat
;;can you do better?
to unknown
set num-unknown-games num-unknown-games + 1
set partner-defected? item ([who] of partner) partner-history
ifelse (partner-defected?) [
set defect-now? true
] [
set defect-now? false
]
end
;;defaults to tit-for-tat-history-update
;;can you do better?
to unknown-history-update
set partner-history
(replace-item ([who] of partner) partner-history partner-defected?)
end
;;;;;;;;;;;;;;;;;;;;;;;;;
;;;Plotting Procedures;;;
;;;;;;;;;;;;;;;;;;;;;;;;;
;;calculate the total scores of each strategy
to do-scoring
set random-score (calc-score "random" num-random)
set cooperate-score (calc-score "cooperate" num-cooperate)
set defect-score (calc-score "defect" num-defect)
set tit-for-tat-score (calc-score "tit-for-tat" num-tit-for-tat)
set unforgiving-score (calc-score "unforgiving" num-unforgiving)
set unknown-score (calc-score "unknown" num-unknown)
end
;; returns the total score for a strategy if any turtles exist that are playing it
to-report calc-score [strategy-type num-with-strategy]
ifelse num-with-strategy > 0 [
report (sum [ score ] of (turtles with [ strategy = strategy-type ]))
] [
report 0
]
end
; Copyright 2002 Uri Wilensky.
; See Info tab for full copyright and license.Завдання: додати інші стратегії поведінки у багатокроковій дилемі ув’язненого. Провести моделювання попарних популяцій та зробити висновок про “найкращу” стратегію.
Soft Grudger - кооперація поки опонент не почав обманювати, потім (щодо цього опонента запускаємо послідовність d,d,d,d,c,c). Потім знову приймаємо рішення.
Gradual - кооперація поки опонент не почав обманювати, потім обманюємо кількість раз, що дорівнює загальній кількості обманів опонента (проти даного агента). Потім два раза кооперуємось.
Pavlov / Random (repeat last choice if good outcome and Random) - Якщо в останньому раунді набрано 5 або 3 то повторюємо останній вибір. Якщо ні - перемикаємось на протилежний. Іноді робимо випадковий вибір.
True Peace Maker (hybrid of Tit For Tat and Tit For Two Tats with Random Co-operation) - кооперуватись поки опонент не обмане два раунди поспіль, потім обманути раз і повернутись до кооперації. Іноді робити випадковий вибір.