BPFL/plot.py at main · amirrezaskh/BPFL · 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

import matplotlib.pyplot as plt
import json
import numpy as np
import seaborn as sns

sns.set_style("whitegrid")

plt.rcParams.update({
    'font.size': 14,
    'axes.labelsize': 16,
    'axes.titlesize': 18,
    'legend.fontsize': 12,
    'figure.figsize': (10, 6),
    'lines.linewidth': 2,
    'lines.markersize': 8
})

# Load the JSON results from file (adjust the filename if needed)
with open('./nodes/results/res.json', 'r') as f:
    data = json.load(f)

# Initialize lists for overall round metrics
rounds = []
global_losses = []
new_model_prices = []

# Initialize dictionaries to store per-wallet metrics
wallets = {}  # Keys will be wallet IDs, values will be dicts with lists for each metric

# Process each round in the JSON data
for round_data in data:
    rounds.append(round_data['round'])
    global_losses.append(round_data['g_model_loss'])
    new_model_prices.append(round_data['new_model_price'])

    # Iterate over each submit (client result) in this round
    for submit in round_data['submits']:
        wid = submit['walletId']
        if wid not in wallets:
            wallets[wid] = {
                'loss': [],
                'delta_local_loss': [],
                'delta_gap': [],
                'contribution': [],
                'reward': []
            }
        wallets[wid]['loss'].append(submit['loss'])
        wallets[wid]['delta_local_loss'].append(submit['delta_local_loss'])
        wallets[wid]['delta_gap'].append(submit['delta_gap'])
        wallets[wid]['contribution'].append(submit['contribution'])
        wallets[wid]['reward'].append(submit['reward'])

# Plot 1: Global Model Loss and New Model Price vs. Rounds
fig, axs = plt.subplots(1, 2, figsize=(14, 6))
axs[0].plot(rounds, global_losses, marker='o', color='navy')
axs[0].set_xlabel('Round')
axs[0].set_ylabel('Global Model Loss')
axs[0].set_title('Global Model Loss Over Rounds')
axs[0].grid(True)

axs[1].plot(rounds, new_model_prices, marker='s', color='darkorange')
axs[1].set_xlabel('Round')
axs[1].set_ylabel('New Model Price')
axs[1].set_title('New Model Price Over Rounds')
axs[1].grid(True)

plt.tight_layout()
plt.savefig('./figures/global_metrics.png', dpi=300)

# Plot 2: Contributions for Each Wallet
plt.figure()
for wid, metrics in wallets.items():
    plt.plot(rounds, metrics['contribution'], marker='x', label=f'{wid}')
plt.xlabel('Round')
plt.ylabel('Contribution')
plt.title('Wallet Contributions Over Rounds')
plt.legend(title='Wallet ID')
plt.tight_layout()
plt.savefig('./figures/wallet_contributions.png', dpi=300)

# Plot 3: Rewards for Each Wallet
plt.figure()
for wid, metrics in wallets.items():
    plt.plot(rounds, metrics['reward'], marker='^', label=f'{wid}')
plt.xlabel('Round')
plt.ylabel('Reward')
plt.title('Wallet Rewards Over Rounds')
plt.legend(title='Wallet ID')
plt.tight_layout()
plt.savefig('./figures/wallet_rewards.png', dpi=300)

# Optional Plot 4: Local Loss for Each Wallet
plt.figure()
for wid, metrics in wallets.items():
    plt.plot(rounds, metrics['loss'], marker='o', label=f'{wid}')
plt.xlabel('Round')
plt.ylabel('Local Loss')
plt.title('Wallet Local Loss Over Rounds')
plt.legend(title='Wallet ID')
plt.tight_layout()
plt.savefig('./figures/wallet_local_loss.png', dpi=300)

# Optional Plot 5: Delta Local Loss per Wallet
plt.figure()
for wid, metrics in wallets.items():
    plt.plot(rounds[1:], metrics['delta_local_loss'][1:], marker='o', label=f'{wid}')
plt.xlabel('Round')
plt.ylabel('Delta Local Loss')
plt.title('Delta Local Loss Over Rounds')
plt.legend(title='Wallet ID')
plt.tight_layout()
plt.savefig('./figures/delta_local_loss.png', dpi=300)

# Optional Plot 6: Delta Gap per Wallet
plt.figure()
for wid, metrics in wallets.items():
    plt.plot(rounds, metrics['delta_gap'], marker='o', label=f'{wid}')
plt.xlabel('Round')
plt.ylabel('Delta Gap')
plt.title('Delta Gap Over Rounds')
plt.legend(title='Wallet ID')
plt.tight_layout()
plt.savefig('./figures/delta_gap.png', dpi=300)