CAPRI2/plot_gaze_behavior_data_model.py at main · CognitiveModeling/CAPRI2 · GitHub

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
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
# Plot the infant data and model predictions

import numpy as np
import matplotlib.pyplot as plt
#import tikzplotlib

# actual data (infant data: Hand: Adam & Elsner (2020) PlosOne, https://doi.org/10.1371/journal.pone.0240165,
# Claw: Adam et al. (2021) Frontiers, https://doi.org/10.3389/fpsyg.2021.695550)

# 	                6 mo Mean	6 mo SE	7 mo Mean	7 mo SE	11 mo Mean	11 mo SE	18 mo Mean	18 mo SE
# hand with effect	-40,91	72	252,09	55	434,6	51
# hand w/o effect	6,53	64	9,92	80	352,38	68
# claw with effect			-2,85	83,7	265,8	92,8	241,2	85,5
# claw w/o effect			-392,2	68,5	-20,7	58,6	221,06	74,8
data = np.array([[-40.91, 72, 252.09, 55, 434.6, 51, np.nan, np.nan],
				 [6.53, 64, 9.92, 80, 352.38, 68, np.nan, np.nan],
				 [np.nan, np.nan, -2.85, 83.7, 265.8, 92.8, 241.2, 85.5],
				 [np.nan, np.nan, -392.2, 68.5, -20.7, 58.6, 221.06, 74.8]])
print(data)

# Comparison plot, with infant data and model

# (copied from other script)
def find_t_look_policy(data):
    for t in range(len(data)):#(270):
        if data[t, policy_t] == 1:
            return t
    return 270
def find_t_look(data):
    for t in range(len(data)):#(270):
        if data[t, policy_t] == 1 or (data[t, event_t] == 3 and data[t, policy_t] == 0):
            return t
    return 270

def normalize_t(data, t_look):
    return  (270 - t_look)/100

def normalize_t_to_one(data, t_look):
    return  (270 - t_look)/270 #(todo: ??)


# models
# - with and without action effects
# - with or without shape adaptation

test_name = "res_tau_2_sim"
folder_name_nonad_full = "Experiments/ResAblationTimeHorizon-test_old_0_4v.1.1-3-adashape-t0-0.4-t34-0-1"
folder_name_adapt_full = "Experiments/ResAblationTimeHorizon-test_old_0_4v.1.1-3-adashape-t0-0.4-t34-1"
folder_name_nonad_cut = "Experiments/ResAblationTimeHorizon-test_old_0_4v.1.1-3-adashape-t0-0.4-t35-c-0"
folder_name_adapt_cut = "Experiments/ResAblationTimeHorizon-test_old_0_4v.1.1-3-adashape-t0-0.4-t35-c-2"
model_folders = [folder_name_nonad_full, folder_name_adapt_full, folder_name_nonad_cut, folder_name_adapt_cut]

# either with or without shape adaptation
model_folders = [folder_name_adapt_full, folder_name_adapt_cut]

epochs = [2, 4, 13, 26]

runs = range(12)
num_runs = len(runs)
sims = range(20)
num_sims = len(sims)

# Processing
a_shapes = ["hand","claw"]
a_shapes = [0.4, 0.8]
event_t = 1 # index of e(t)
policy_t = 2 # index of pi(t)
looking_ts_all = np.zeros((len(model_folders), num_sims, len(a_shapes), len(epochs), num_runs))
looking_ts_all_policy = np.zeros((len(model_folders), num_sims, len(a_shapes), len(epochs), num_runs))
for model, foldername in enumerate(model_folders):
	for sim_idx, sim in enumerate(sims):
		for agent, a_shape in enumerate(a_shapes):
			for ep, epoch in enumerate(epochs):
				for run_idx, run in enumerate(runs):
					log_file_name = foldername + "/" + test_name + str(sim) + "/log_files/"
					filename = log_file_name + "res_tau_2_sim" + str(sim) + "_epoch" + str(epoch) + "_" + str(
						a_shape) + "_run" + str(run) + ".txt"
					data_0 = np.loadtxt(filename, dtype='float64', skiprows = 1, delimiter= ', ')
					t_look = find_t_look(data_0)
					t_look_policy = find_t_look_policy(data_0)
					looking_ts_all[model, sim_idx, agent, ep, run_idx] = normalize_t(data, t_look)
					looking_ts_all_policy[model, sim_idx, agent, ep, run_idx] = normalize_t(data, t_look_policy)
looking_ts_all_mean = np.mean(np.mean(looking_ts_all_policy, axis= 3), axis= 1)
looking_ts_hand_policy = looking_ts_all_policy[:,:,0,:,:]
looking_ts_claw_policy = looking_ts_all_policy[:,:,1,:,:]
looking_ts_hand_mean = np.mean(np.mean(looking_ts_hand_policy, axis=3), axis = 1)
looking_ts_hand_sd = np.std(np.mean(looking_ts_hand_policy, axis=3), axis=1)
looking_ts_claw_mean = np.mean(np.mean(looking_ts_claw_policy, axis=3), axis=1)
looking_ts_claw_sd = np.std(np.mean(looking_ts_claw_policy, axis=3), axis=1)

# (hand and claw, each, for two models - with or without adaptation/action effects)
looking_ts_all_1 = np.zeros((num_sims, len(model_folders)*len(a_shapes), len(epochs), num_runs))
looking_ts_all_policy_1 = np.zeros((num_sims, len(model_folders)*len(a_shapes), len(epochs), num_runs))
for sim_idx, sim in enumerate(sims):
	i = -1
	for agent, a_shape in enumerate(a_shapes):
		for model, foldername in enumerate(model_folders):
			i = i+1
			for ep, epoch in enumerate(epochs):
				for run_idx, run in enumerate(runs):#range(num_runs):
					log_file_name = foldername + "/" + test_name + str(sim) + "/log_files/"
					filename = log_file_name + "res_tau_2_sim" + str(sim) + "_epoch" + str(epoch) + "_" + str(
						a_shape) + "_run" + str(run) + ".txt"
					data_0 = np.loadtxt(filename, dtype='float64', skiprows = 1, delimiter= ', ')
					t_look = find_t_look(data_0)
					t_look_policy = find_t_look_policy(data_0)
					looking_ts_all_1[sim_idx, i, ep, run_idx] = normalize_t(data, t_look)
					looking_ts_all_policy_1[sim_idx, i, ep, run_idx] = normalize_t(data, t_look_policy)

looking_ts_all_mean = np.mean(np.mean(looking_ts_all_policy_1, axis= 3), axis=0)
print(looking_ts_all_mean)
print(looking_ts_all_mean.shape)

# Generate the plot

fig, ax = plt.subplots()
ax.plot([0,1,2], data[0,0:6:2], label='hand with effect', marker='o')
ax.plot([0,1,2], data[1,0:6:2], label='hand w/o effect', marker='o')
ax.plot([1,2,3], data[2,2:8:2], label='claw with effect', marker='o')
ax.plot([1,2,3], data[3,2:8:2], label='claw w/o effect', marker='o')
ax.set_ylabel('Gaze arrival times (ms)')
ax.set_xticks([0,1,2,3], ['6 mo', '7 mo', '11 mo', '18 mo'])
#ax.grid(True, which='major', axis='y')
fig.legend(loc='lower right', bbox_to_anchor=(0.89,0.12))
ax2 = ax.twinx()
for i in range(len(model_folders)*len(a_shapes)):
	ax2.plot([0, 1, 2, 3], looking_ts_all_mean[i], marker='d', linestyle='--')
ax2.set_ylim([0.7,2.7]) #(?)
ax2.set_yticks([1.25, 1.7, 2.25, 2.7], ('e_transport', '', 'e_reach', ''))
#plt.title('Gaze Arrival Times for Agents/Actions')
plt.savefig("gaze_arrival_times_comparison_infants_n4_epochs" + str(epochs) + ".png", bbox_inches="tight")
#tikzplotlib.save("gaze_arrival_times_comparison_infants_epochs" + str(epochs) + ".tex")
plt.show()