-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathconsts_dict.py
64 lines (55 loc) · 3.55 KB
/
consts_dict.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
#########################################
## The default set of parameters for a ##
## nest transfer experiment simulation ##
## (c) Artem Pashchinskiy, UCLA, 2019 ##
#########################################
from math import pi
pars = {
#### General parameters
'ARENA' : 'c0d0', # which configuration of the arena to use; see getTypeFile in Arena.py for details
'NUM' : 15, # the numer of ants in simulation
'ITER' : 3000, # length of simulations in iterations
# parametrization was based on frames in 30 fps videos
# which yields a conversion rate of 1800 iterations <-> 1 min real time
'SIMSPEED' : 500, # 1 / frequency of snapshots taken for trajectories - used only if trajectories are recorded
#### Parameters controlling walking patterns
'ANT_STEP' : 2.3, # mean of the step distribution (mu)
'ANT_STEP_VAR' : 12.4, # variance of the step distribution (sigma)
'ANT_STEP_MAX' : 6.44, # maximal cut-off for the length of steps (maxstep)
'RANDANGLE' : (-30, 30), # DEPRECIATED
'RANDFREQ' : 1, # DEPRECIATED
'ANGLEEXPMEAN':pi/1.5, # variance of normal distribution used in defining turtousity (sigma')
'WALL' : 'deldir', # 'deldir' - change direction until avaliable step is found;
# 'delstep' - decrease step size until can move in the original direction;
'MAXTRY' : 5, # how many attempts at avoiding an obstacle an agent will take in one timestamp
#### Parameters controling bias
'BIAS_C' : 'p', # type of bias: 'l' for linerar, 'c' for constant, 'p' for parabolic
'BIAS_PAR' : 2, # multiplier for length of step from general bias (B)
'PARAB_ZERO' : 4, # part of the field (1/x) that lies to the left of zero of parabolic bias in x direction
# (e.g. 4 stands for the bias vertex being placed at the distance of 25% of the arena length from the left edge)
#### Parameters controlling initial placement
'INIT_DISTR' : 'u', # initial distribution of ants in x-axis. 'u' - uniform
# 'ls' - right-skewed, 'rs' - left-skewed, 'e' - expontential
'ORIGIN' : (1470, 540), # location of spawning region center
# usually overwritten automatically when a specific arena is loaded
# see getFieldType in Arena.py for more details
'ESCALE' : 125, # parameter of truncated exponential distr of X-coordinate of initial placement
'SHIFT' : 300, # DEPRECIATED
###################################################################################
#### Parameters controlling interaction-based activation (not used in the research)
'ACTIVATED_BIAS_PAR' : 2, # multiplier for length of step for the activated state
'EXCITED_BIAS_PAR' : 5, # multiplier for additional length of step from excited state
#### Parameters controlling interactions (not used in the research)
'EXCITMENT_TRESHOLD' : 5, # from what excitment level (in EP - Exitement Points) new bias kicks-in
'ONE_INTER_INCR' : 50, # how much EP is added from one interaction with activated ant
'ACTIVATION_ZONE' : 500, # part of the field (1/x) around parabola zero X where ants have a chance to be activated
'ACTIVATION_CHANCE' : 10, # chance (1/x) ant that is in the activation zone actually gets activated
'DEACTIVATION_DELAY' : 900, # time (in iterationss) till activated ant gets deactivated again
'ANT_SIZE_X' : 16,
'ANT_SIZE_Y' : 8,
'INTER_RAD' : 8, # radius of interaction-determining poximity
}
#### Constsnts using other parameters
#DEACTIVATION_DELAY = pars['ITER'] // 30,
# for how many iterations AFTER reaching the center of mass an actiated ant remains activated
#pars['DEACTIVATION_DELAY'] = DEACTIVATION_DELAY[0]