-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathutils.py
172 lines (134 loc) · 6.34 KB
/
utils.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
import os
import torch
import random
import numpy as np
import pandas as pd
import scipy.sparse as sp
from sklearn.metrics import accuracy_score
from sklearn.metrics.pairwise import cosine_similarity
from gnn_cp.cp.cp_manager import CPManager
import gnn_cp.models.graph_models as graph_models
from gnn_cp.data.data_manager import GraphDataManager
from gnn_cp.models.model_manager import GraphModelManager
device = "cuda" if torch.cuda.is_available() else "cpu"
print(f"Torch Graph Models are running on {device}")
def make_dataset_instances(
dataset_manager: GraphDataManager,
dataset_key,
dataset,
model_class_name,
models_config,
splits_config,
models_cache_dir):
instances = dataset_manager.load_splits(dataset_key, dataset, splits_config)
print("Dataset Loaded Successfully!")
print("====================================")
print(f"Loading Models {model_class_name}")
model_class = getattr(graph_models, model_class_name)
for instance_idx, instance in enumerate(instances):
model_manager = GraphModelManager(
model_class_name=model_class_name,
model_class=model_class,
models_config=models_config,
dataset=dataset,
checkpoint_address=models_cache_dir,
model_name=f"{dataset_key}-ins{instance_idx}-{model_class.__name__}")
model_manager.load_model(dataset,instance)
model_manager.model = model_manager.model.to(device)
y_pred = model_manager.predict(
dataset, test_idx=instance['test_idx'], return_embeddings=False
)
y_embeddings = model_manager.predict(
dataset, return_embeddings=True
)
accuracy = accuracy_score(
y_true=dataset.y[instance['test_idx']].cpu().numpy(),
y_pred=y_pred.cpu().numpy(),
)
instance.update({"model": model_manager, "accuracy": accuracy, "embeddings": y_embeddings})
print(
f"Accuracy: {np.mean([instance['accuracy'] for instance in instances])} ± {np.std([instance['accuracy'] for instance in instances])}"
)
return instances
def get_overall_cp_result(
dataset_manager: GraphDataManager,
dataset,
instances,
selected_coverage,
dataset_name,
edge_index_initial,
adj_knn):
tune_fraction, calib_fraction, notune_calib_fraction = compute_tune_calib_fraction(instances[0])
cp_manager = CPManager(dataset=dataset,
coverage_val=selected_coverage,
tune_fraction=tune_fraction,
calib_fraction=calib_fraction,
notune_calib_fraction=notune_calib_fraction,
dataset_name=dataset_name,
edge_index=edge_index_initial,
test_idx=None,
adj_knn=adj_knn)
instance_results = []
for instance in instances:
embeddings = instance["embeddings"]
tune_idx, test_idx = tune_truetest_split(dataset_manager, instance["test_idx"], dataset, tune_fraction)
cp_keys = ['APS']
cp_keys.append('DAPS-APS')
cp_keys.append('SNAPS-APS')
cp_keys.append('TSNAPS-APS')
res = cp_manager.get_all_cp_results(cp_keys, embeddings, tune_idx, test_idx, instance["test_idx"])
instance_results.append(res['mean'])
instance_results = pd.concat(instance_results, axis=0, keys=[idx for idx in range(len(instances))])
instance_mean_results = instance_results.reset_index().rename(
columns={"level_0": "instance", "level_1": "method"})
return instance_mean_results
def compute_tune_calib_fraction(instance, max_calib_num=1000):
tune_num = min(instance["train_idx"].shape[0], 500)
tune_fraction = tune_num / instance["test_idx"].shape[0]
calib_num = min(max_calib_num, int(instance["test_idx"].shape[0] / 2))
calib_fraction = (calib_num - tune_num) / (instance["test_idx"].shape[0] - tune_num)
notune_calib_num = min(max_calib_num, int(instance["test_idx"].shape[0] / 2))
notune_calib_fraction = notune_calib_num / instance["test_idx"].shape[0]
return tune_fraction, calib_fraction, notune_calib_fraction
def tune_truetest_split(dataset_manager, test_idx, dataset, tuning_fraction):
te_idx, tu_idx, _, _ = dataset_manager.train_test_split(test_idx, dataset.y[test_idx], training_fraction=tuning_fraction)
return te_idx, tu_idx
def print_results(models_results, model_classes):
models_results = pd.concat(models_results, axis=0, keys=model_classes)
result = models_results.reset_index().rename(columns={"level_0": "model"})
average_result = result.groupby(
["model","method"], sort=False).mean().reset_index().drop(columns=["level_1", "instance"])
print(average_result.to_markdown())
def sparse_mx_to_torch_sparse_tensor(sparse_mx):
"""Convert a scipy sparse matrix to a torch sparse tensor."""
sparse_mx = sparse_mx.tocoo().astype(np.float32)
indices = torch.from_numpy(
np.vstack((sparse_mx.row, sparse_mx.col)).astype(np.int64))
values = torch.from_numpy(sparse_mx.data)
shape = torch.Size(sparse_mx.shape)
return torch.sparse.FloatTensor(indices, values, shape)
def compute_adj_knn(dataset_name, features, k=20):
if not os.path.exists('./results/adj_knn/{}_adj_knn_{}'.format(dataset_name, k)):
features = np.copy(features.cpu())
features[features != 0] = 1
sims = cosine_similarity(features)
sims[(np.arange(len(sims)), np.arange(len(sims)))] = 0
for i in range(len(sims)):
indices_argsort = np.argsort(sims[i])
sims[i, indices_argsort[:-k]] = 0
A_feat = sp.coo_matrix(sims)
row_sum = np.array(A_feat.sum(1))
d_inv = np.power(row_sum, -1.0).flatten()
d_inv[np.isinf(d_inv)] = 0.
d_mat_inv = sp.diags(d_inv)
A_feat = d_mat_inv.dot(A_feat).tocoo()
adj_knn_st = sparse_mx_to_torch_sparse_tensor(A_feat).float()
torch.save(adj_knn_st, './results/adj_knn/{}_adj_knn_{}.pt'.format(dataset_name, k))
else:
adj_knn_st = torch.load('./results/adj_knn/{}_adj_knn_{}.pt'.format(dataset_name, k))
return adj_knn_st
def set_seed(seed):
random.seed(seed)
np.random.seed(seed)
torch.manual_seed(seed)
torch.cuda.manual_seed_all(seed)