-
Notifications
You must be signed in to change notification settings - Fork 0
/
plot_comparative_meta-testing.py
175 lines (121 loc) · 6.38 KB
/
plot_comparative_meta-testing.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
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
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
import json
import pandas as pd
import os
import fnmatch
import matplotlib.pyplot as plt
from pathlib import Path
import numpy as np
def find_files(directory, file_name, dataset):
matches = []
for root, dirnames, filenames in os.walk(directory):
for filename in fnmatch.filter(filenames, file_name):
matches.append(os.path.join(root, filename))
return matches
def plot_data(data_plot, ylabel, ax, i, dsc):
colors = ['blueviolet' ,'orange', 'lime', 'black', 'blue','lightgray', 'blueviolet' ]
data_plot = data_plot.reset_index()
grouped = data_plot.groupby('Name')
color_id = 0
for name, group in grouped:
if name == 'baseline':
name = 'batch'
elif name== 'oml':
name = name.upper()
else:
if name == 'maml':
name = 'MAML-Rep'
ax.errorbar(group['Nr_classes'], group[ylabel], yerr=group[ylabel+'_std'], label=name, markersize=1, linestyle='-', color=colors[color_id], linewidth=0.8, elinewidth=0.3)
color_id += 1
ax.set_xlabel('Number of Classes',fontsize=12)
ax.set_ylabel('Scores', fontsize=12)
if ylabel == 'Accuracy':
title = 'F1 micro'
elif ylabel == 'F1 score':
title = 'F1 macro'
else:
title = 'Precision'
if i in [1,2,4,5]:
ax.yaxis.set_visible(False)
ax.set_ylim(0, 1.1) # Adjust ylim as needed
plt.text(0.95, 0.95,dsc, ha='right', va='top', transform=ax.transAxes,fontsize=12)
ax.set_xticks(data_plot['Nr_classes'].unique().tolist())
if i in range(0,3):
ax.set_title(title)
ax.xaxis.set_visible(False)
def main():
ref_datasets = ['dsads','ucihar', 'pamap2', 'hapt']
# path to results to plot
main_path ='/home/.../'
file_name = 'metadata.json'
# directory to save plots
ref = 'comparative/'
for dataset in ref_datasets:
path = main_path + dataset
result_path = find_files(path, file_name, dataset)
columns = ['Name', 'Nr_classes', 'Id', 'Accuracy', 'F1 score', 'F1 score weighted', 'Precision', 'Accuracy_std', 'F1 score_std', 'F1 score weighted_std', 'Precision_std' ]
# all
keys = ['Train average stats','Test average stats','Train average stats base','Test average stats base']
keys_id = [ 'Train', 'Test','TrainBase', 'TestBase']
df = pd.DataFrame(columns=columns)
for path in result_path:
with open(path, 'r') as f:
obj = json.load(f)
data = obj.get('params')
schedule = data['schedule']
name = data['name']
aug = data.get('augmentation_ref')
for nr_classes in schedule:
idx = 0
for key in keys_id:
name = obj.get('name')
data = obj.get('results').get(keys[idx] + ' ' + str(nr_classes))
instance ={ 'Name' : name,
'Nr_classes': nr_classes,
'Id': key,
'Aug': aug,
'Accuracy': data['Accuracy'],
'F1 score': data['F1-score macro'],
'F1 score weighted': data['F1-score weighted'],
'Precision': data['Macro precision'],
'Accuracy_std': data['Accuracy std'],
'F1 score_std': data['F1-score macro std'],
'F1 score weighted_std': data['F1-score weighted std'],
'Precision_std': data['Macro precision std']}
new_df = pd.DataFrame([instance])
df = pd.concat([df,new_df], ignore_index=True)
idx += 1
for aug in np.unique(df['Aug']):
directory = Path(main_path + ref + dataset + '/' + aug)
if not directory.exists():
directory.mkdir(parents=True, exist_ok=True)
fig, axs = plt.subplots(nrows=2, ncols=3, figsize=(12, 8))
fig.subplots_adjust(wspace=0.02, hspace=0.02)
row = 0
for i in range(0,len(keys_id)):
id_plot = keys_id[i]
df_plot = df.loc[(df['Id'] ==id_plot) & (df['Aug']==aug)]
# Plotting accuracy
data_plot_accuracy = df_plot.groupby(['Name', 'Nr_classes'])[['Accuracy', 'Accuracy_std']].mean()
data_plot_accuracy = data_plot_accuracy.sort_values(by='Name')
plot_data(data_plot_accuracy, 'Accuracy', axs[i][0], i+row, id_plot)
# Plotting F1 score
data_plot_f1 = df_plot.groupby(['Name', 'Nr_classes'])[['F1 score', 'F1 score_std']].mean()
data_plot_f1 = data_plot_f1.sort_values(by='Name')
plot_data(data_plot_f1, 'F1 score', axs[i][1], i+1+row, id_plot)
# Plotting Precision
data_plot_precision = df_plot.groupby(['Name', 'Nr_classes'])[['Precision', 'Precision_std']].mean()
data_plot_precision = data_plot_precision.sort_values(by='Name')
plot_data(data_plot_precision, 'Precision', axs[i][2],i+2+row, id_plot)
row = 2
# Add legend without border
# Combine legends
handles, labels = axs[0, 0].get_legend_handles_labels()
fig.legend(handles, labels, loc='lower center', bbox_to_anchor=(0.5, -0.001), frameon=False, fancybox=True, shadow=True, ncol=len(labels),fontsize = 14)
plt.suptitle(dataset.upper(),x=0.5, y=0.94, fontsize = 14)
plt.savefig(str(directory) + "/" + dataset+'_'+aug +'600', dpi=600, bbox_inches='tight')
plt.savefig(str(directory) + "/" + dataset+'_'+aug +'.pdf', bbox_inches='tight')
plt.savefig(str(directory) + "/" + dataset+'_'+aug +'.svg', bbox_inches='tight')
if __name__ == '__main__':
main()