@@ -0,0 +1,76 @@ | |||
MGCN: Semi-supervised Classification in Multi-layer Graphs with Graph Convolutional Networks | |||
==== | |||
Here is the code for node embedding in multi-layer networks with attributes written in Pytorch. | |||
Ghorbani et.al. "MGCN: Semi-supervised Classification in Multi-layer Graphs with Graph Convolutional Networks" [1] | |||
Usage | |||
------------ | |||
The main file is "train.py". It contains the running code for infra dataset with default parameters. | |||
```python train.py``` | |||
Input Data | |||
------------ | |||
You should create a folder in "data" folder, named "dataset_str" in the code. For example if your dataset_str is "infra", you should place files in the path "data/infra/". The folder contains the following information: | |||
1) The adjacency list of within layer edges for every layer. Node numbering should be started from 0. The name of the files containing the within layer adjacency list should follow the pattern "dataset_name.adj#NumberOfLayer". | |||
For example if your dataset name is "infra" with three layers, you should have "infra.adj0", "infra.adj1" and "infra.adj2" files that every one of them contains adjacency list like follow: | |||
``` | |||
0 1 | |||
0 2 | |||
0 3 | |||
0 4 | |||
``` | |||
2) The adjacency list of between layer edges for every pair of layers (if available). The name of the files containing the between layer adjacency list should follow the pattern "dataset_name.bet#NumberOfLayer1_#NumberOfLayer2". | |||
For example if your dataset name is "infra" with three layers, you should have "infra.bet0_1", "infra.bet0_2" and "infra.bet1_2" (One file for every pair of layers is enough) | |||
3) Features and labels of every node in every layer. If features aren't available identity matrix should be used. The name of the files containing the features and labels should follow the pattern "dataset_name.feat#NumberOfLayer" | |||
For example if your dataset name is "infra" with three layers, you should have "infra.feat0", "infra.feat1", "infra.feat2". Every line of file includes the node number, node features, node label. For example for "infra" dataset, if the first layer have 5 node belongs to 2 classes, it should be like follow: | |||
``` | |||
NodeNumber Features Label | |||
0 1 0 0 0 0 1 | |||
1 0 1 0 0 0 1 | |||
2 0 0 1 0 0 2 | |||
3 0 0 0 1 0 2 | |||
4 0 0 0 0 1 2 | |||
``` | |||
Parameters | |||
------------ | |||
Parameters of the code are all in the list format for testing different combination of configs. | |||
Parameters are as follow: | |||
- dataset_str: Dataset name | |||
- adj_weights: It is a list for assigning different weights to input layers. If you leave it empty, all layers will have equal weihts. [[1,1,1]] includes one config with equal weights. | |||
- wlambdas: It is the weight of reconstruction loss. | |||
- hidden_structures: A list contains different structures for hidden spaces in every layer. For example [[[32],[32],[32]]] contains one config for hidden spaces of every layer which in all of them, the dimension of embedding for all layers are 32. | |||
- lrs: A list contains learning rate(s). | |||
- test_sizes: A list contains the test size(s). | |||
Note | |||
------------ | |||
Thanks to Thomas Kipf. The code is written based on the "Graph Convolutional Networks in PyTorch" [2]. | |||
Bug Report | |||
------------ | |||
If you find a bug, please send a bug report to [email protected] including if necessary the input file and the parameters that caused the bug. | |||
You can also send me any comment or suggestion about the program. | |||
References | |||
------------ | |||
[1] [Ghorbani et.al., MGCN: Semi-supervised Classification in Multi-layer Graphs with Graph Convolutional Networks, 2019](https://arxiv.org/pdf/1811.08800) | |||
[2] [Kipf & Welling, Semi-Supervised Classification with Graph Convolutional Networks, 2016](https://arxiv.org/abs/1609.02907) | |||
Cite | |||
------------ | |||
Please cite our paper if you use this code in your own work: | |||
``` | |||
@article{ghorbani2018multi, | |||
title={Multi-layered Graph Embedding with Graph Convolution Networks}, | |||
author={Ghorbani, Mahsa and Baghshah, Mahdieh Soleymani and Rabiee, Hamid R}, | |||
journal={arXiv preprint arXiv:1811.08800}, | |||
year={2018} | |||
} | |||
``` |
@@ -0,0 +1,80 @@ | |||
import math | |||
import torch | |||
from torch.nn.parameter import Parameter | |||
from torch.nn.modules.module import Module | |||
class GraphConvolution(Module): | |||
""" | |||
Simple GCN layer, similar to https://arxiv.org/abs/1609.02907 | |||
""" | |||
def __init__(self, in_features, out_features, bias=True): | |||
super(GraphConvolution, self).__init__() | |||
self.in_features = in_features | |||
self.out_features = out_features | |||
self.weight = Parameter(torch.FloatTensor(in_features, out_features)) | |||
if bias: | |||
self.bias = Parameter(torch.FloatTensor(out_features)) | |||
else: | |||
self.register_parameter('bias', None) | |||
self.reset_parameters() | |||
def reset_parameters(self): | |||
stdv = 1. / math.sqrt(self.weight.size(1)) | |||
self.weight.data.uniform_(-stdv, stdv) | |||
if self.bias is not None: | |||
self.bias.data.uniform_(-stdv, stdv) | |||
def forward(self, input, adj): | |||
support = torch.spmm(input, self.weight) | |||
output = torch.spmm(adj, support) | |||
if self.bias is not None: | |||
return output + self.bias | |||
else: | |||
return output | |||
def __repr__(self): | |||
return self.__class__.__name__ + ' (' \ | |||
+ str(self.in_features) + ' -> ' \ | |||
+ str(self.out_features) + ')' | |||
class CrossLayer(Module): | |||
""" | |||
MultiLayer | |||
""" | |||
def __init__(self, L1_dim, L2_dim, bias=True, bet_weight=True): | |||
super(CrossLayer, self).__init__() | |||
self.L1_dim = L1_dim | |||
self.L2_dim = L2_dim | |||
self.bet_weight = bet_weight | |||
self.weight = Parameter(torch.FloatTensor(L1_dim, L2_dim)) | |||
if bias: | |||
self.bias = Parameter(torch.FloatTensor(L2_dim)) | |||
else: | |||
self.register_parameter('bias', None) | |||
self.reset_parameters() | |||
def reset_parameters(self): | |||
stdv = 1. / math.sqrt(self.weight.size(1)) | |||
self.weight.data.uniform_(-stdv, stdv) | |||
if self.bias is not None: | |||
self.bias.data.uniform_(-stdv, stdv) | |||
def forward(self, L1_features, L2_features): | |||
if self.bet_weight: | |||
temp = torch.mm(L1_features, self.weight) | |||
output = torch.mm(temp, torch.t(L2_features)) | |||
if self.bias is not None: | |||
output = output + self.bias | |||
else: | |||
output = torch.mm(L1_features, torch.t(L2_features)) | |||
return output | |||
def __repr__(self): | |||
return self.__class__.__name__ + ' (' \ | |||
+ str(self.L1_dim) + ' -> ' \ | |||
+ str(self.L2_dim) + ')' |
@@ -0,0 +1,113 @@ | |||
import torch | |||
import torch.nn as nn | |||
import torch.nn.functional as F | |||
from layers import GraphConvolution, CrossLayer | |||
from utils import between_index | |||
class GCN(nn.Module): | |||
def __init__(self, nfeat, nhid, nclass, dropout): | |||
super(GCN, self).__init__() | |||
self.nclass = nclass | |||
layers = [] | |||
layers.append(GraphConvolution(nfeat, nhid[0])) | |||
for i in range(len(nhid)-1): | |||
layers.append(GraphConvolution(nhid[i], nhid[i+1])) | |||
if nclass > 1: | |||
layers.append(GraphConvolution(nhid[-1], nclass)) | |||
self.gc = nn.ModuleList(layers) | |||
self.dropout = dropout | |||
def forward(self, x, adj): | |||
end_layer = len(self.gc)-1 if self.nclass > 1 else len(self.gc) | |||
for i in range(end_layer): | |||
x = F.relu(self.gc[i](x, adj)) | |||
x = F.dropout(x, self.dropout, training=self.training) | |||
if self.nclass > 1: | |||
classifier = self.gc[-1](x, adj) | |||
return F.log_softmax(classifier, dim=1), x | |||
else: | |||
return None, x | |||
class CCN(nn.Module): | |||
def __init__(self, n_inputs, inputs_nfeat, inputs_nhid, inputs_nclass, dropout): | |||
super(CCN, self).__init__() | |||
# Features of network | |||
self.n_inputs = n_inputs | |||
self.inputs_nfeat = inputs_nfeat | |||
self.inputs_nhid = inputs_nhid | |||
self.inputs_nclass = inputs_nclass | |||
self.dropout = dropout | |||
# Every single layer | |||
temp_in_layer = [] | |||
temp_bet_layer = [] | |||
for i in range(n_inputs): | |||
temp_in_layer.append(GCN(inputs_nfeat[i], inputs_nhid[i], inputs_nclass[i], dropout)) | |||
self.in_layer = nn.ModuleList(temp_in_layer) | |||
# Between layers | |||
for i in range(n_inputs): | |||
for j in range(i+1,n_inputs): | |||
temp_bet_layer.append(CrossLayer(inputs_nhid[i][-1], inputs_nhid[j][-1], bias=True, bet_weight=False)) | |||
self.bet_layer = nn.ModuleList(temp_bet_layer) | |||
def forward(self, xs, adjs): | |||
classes_labels = [] | |||
bet_layers_output = [] | |||
in_layers_output = [] | |||
features = [] | |||
# Single layer forward to find features and classes | |||
for i in range(self.n_inputs): | |||
class_labels, feature = self.in_layer[i](xs[i], adjs[i]) | |||
classes_labels.append(class_labels) | |||
features.append(feature) | |||
temp = torch.mm(feature, torch.t(feature)) | |||
in_layers_output.append(temp) | |||
# Find between layers with CCN | |||
for i in range(self.n_inputs): | |||
for j in range(i+1,self.n_inputs): | |||
temp = self.bet_layer[between_index(self.n_inputs, i, j)](features[i], features[j]) | |||
bet_layers_output.append(temp) | |||
return classes_labels, bet_layers_output, in_layers_output | |||
class AggCCN(nn.Module): | |||
def __init__(self, n_inputs, inputs_nfeat, inputs_nhid, inputs_nclass, dropout): | |||
super(AggCCN, self).__init__() | |||
# Features of network | |||
self.n_inputs = n_inputs | |||
self.inputs_nfeat = inputs_nfeat | |||
self.inputs_nhid = inputs_nhid | |||
self.inputs_nclass = inputs_nclass | |||
self.dropout = dropout | |||
# Every single layer | |||
temp_in_layer = [] | |||
temp_bet_layer = [] | |||
for i in range(n_inputs): | |||
temp_in_layer.append(GCN(inputs_nfeat[i], inputs_nhid[i], inputs_nclass[i], dropout)) | |||
self.in_layer = nn.ModuleList(temp_in_layer) | |||
def forward(self, xs, adjs): | |||
classes_labels = [] | |||
bet_layers_output = [] | |||
in_layers_output = [] | |||
features = [] | |||
# Single layer forward to find features and classes | |||
for i in range(self.n_inputs): | |||
class_labels, feature = self.in_layer[i](xs[i], adjs[i]) | |||
classes_labels.append(class_labels) | |||
features.append(feature) | |||
temp = torch.mm(feature, torch.t(feature)) | |||
in_layers_output.append(temp) | |||
return classes_labels, in_layers_output |
@@ -0,0 +1,271 @@ | |||
from __future__ import division | |||
from __future__ import print_function | |||
import time | |||
import argparse | |||
import torch | |||
import torch.nn.functional as F | |||
import torch.optim as optim | |||
from sklearn.metrics import f1_score | |||
from utils import load_data, class_accuracy, class_f1, layer_accuracy, layer_f1, dict_to_writer, trains_vals_tests_split | |||
from models import CCN | |||
from tensorboardX import SummaryWriter | |||
# f1 name should be f1_micro or f1_macro | |||
# class_metrics = {'loss': F.nll_loss, 'acc': class_accuracy, 'f1_micro': class_f1, 'f1_macro': class_f1, | |||
# 'all_f1_micro': f1_score, 'all_f1_macro': f1_score} | |||
class_metrics = {'loss': F.nll_loss, 'all_f1_micro': f1_score, 'all_f1_macro': f1_score} | |||
layer_metrics = {'loss': torch.nn.BCEWithLogitsLoss()} | |||
# Preparing the output of classifier for every layer separately (class numbers should be different over layers) | |||
def prepare_output_labels(outputs, labels, idx): | |||
tmp_labels = [] | |||
tmp_outputs = [] | |||
max_label = 0 | |||
for i in range(len(labels)): | |||
tmp_labels.extend(labels[i][idx[i]] + max_label) | |||
tmp_outputs.extend((outputs[i].max(1)[1].type_as(labels[i]) + max_label)[idx[i]]) | |||
max_label = labels[i].max() | |||
return tmp_outputs, tmp_labels | |||
# Calculate classification metrics for every layer | |||
def model_stat_in_layer_class(in_classes_pred, in_labels, idx, metrics): | |||
stats = {} | |||
for i in range(len(in_classes_pred)): | |||
if not in_classes_pred[i] is None: | |||
for metr_name,metr_func in metrics.items(): | |||
if not metr_name in stats: | |||
stats[metr_name] = [] | |||
if metr_name[:2] == 'f1': | |||
try: | |||
stats[metr_name].append(metr_func(in_classes_pred[i][idx[i]], in_labels[i][idx[i]], metr_name.split('f1_')[-1])) | |||
except: | |||
print("Exception in F1 for class" + str(i) + " for classification") | |||
stats[metr_name].append(0) | |||
elif metr_name[:3] == 'all': | |||
continue | |||
else: | |||
stats[metr_name].append(metr_func(in_classes_pred[i][idx[i]], in_labels[i][idx[i]])) | |||
else: | |||
for metr_name, metr_func in metrics.items(): | |||
if not metr_name in stats: | |||
stats[metr_name] = [] | |||
stats[metr_name].append(0) | |||
tmp_in_classes_pred, tmp_labels = prepare_output_labels(in_classes_pred, in_labels, idx) | |||
for metr_name, metr_func in metrics.items(): | |||
if metr_name[:3] == 'all': | |||
stats[metr_name] = metr_func(tmp_in_classes_pred, tmp_labels, average=metr_name.split('f1_')[-1]) | |||
return stats | |||
# Calculate reconstruction metrics for within and between layer edges (based on the input) | |||
def model_stat_struc(preds, reals, metrics, pos_weights, norms, idx=None): | |||
stats = {} | |||
for i in range(len(preds)): | |||
if reals[i] is None: | |||
continue | |||
if 'sparse' in reals[i].type(): | |||
curr_real = reals[i].to_dense() | |||
else: | |||
curr_real = reals[i] | |||
if idx is None: | |||
final_preds = preds[i] | |||
final_real = curr_real | |||
else: | |||
final_preds = preds[i][idx[i][0],idx[i][1]] | |||
final_real = curr_real[idx[i][0],idx[i][1]] | |||
for metr_name,metr_func in metrics.items(): | |||
if not metr_name in stats: | |||
stats[metr_name] = [] | |||
if metr_name == 'f1': | |||
try: | |||
stats[metr_name].append(metr_func(final_preds, final_real, type = metr_name.split('f1_')[-1])) | |||
except: | |||
print("Exception in F1 for layer" + str(i) + " for structure") | |||
stats[metr_name].append(0) | |||
elif metr_name == 'loss': | |||
if args.cuda: | |||
pw = pos_weights[i] * torch.ones(final_real.size()).cuda() | |||
else: | |||
pw = pos_weights[i] * torch.ones(final_real.size()) | |||
stats[metr_name].append(norms[i]*torch.nn.BCEWithLogitsLoss(pos_weight=pw)(final_preds, final_real)) | |||
else: | |||
try: | |||
stats[metr_name].append(metr_func(final_preds, final_real)) | |||
except: | |||
print("Odd problem") | |||
return stats | |||
def train(epoch): | |||
t = time.time() | |||
model.train() | |||
optimizer.zero_grad() | |||
# Run model | |||
t1 = time.time() | |||
classes_pred, bet_layers_pred, in_layers_pred = model(features, adjs) | |||
# Final statistics | |||
stats = dict() | |||
stats['in_class'] = model_stat_in_layer_class(classes_pred, labels, idx_trains, class_metrics) | |||
# Reconstructing all the within and between layer edges without train and test split (for using part of the edges | |||
# for reconstruction, idx is needed) | |||
stats['in_struc'] = model_stat_struc(in_layers_pred, adjs_orig, layer_metrics, adjs_pos_weights, adjs_norms) | |||
stats['bet_struc'] = model_stat_struc(bet_layers_pred, bet_adjs_orig, layer_metrics, bet_pos_weights, bet_norms) | |||
# Write to writer | |||
for name, stat in stats.items(): | |||
dict_to_writer(stat, writer, epoch, name, 'Train') | |||
# Calculate loss | |||
loss_train = 0 | |||
for name, stat in stats.items(): | |||
# Weighted loss | |||
if name == 'in_struc': | |||
loss_train += wlambda * sum([a * b for a, b in zip(stat['loss'], adj_weight)]) | |||
elif name== 'bet_struc': | |||
loss_train += wlambda * sum([a * b for a, b in zip(stat['loss'], adj_weight)]) | |||
else: | |||
loss_train += sum([a * b for a, b in zip(stat['loss'], adj_weight)]) | |||
# Gradients | |||
loss_train.backward() | |||
optimizer.step() | |||
# Validation if needed | |||
if not args.fastmode: | |||
model.eval() | |||
classes_pred, bet_layers_pred, in_layers_pred = model(features, adjs) | |||
stats = dict() | |||
stats['in_class'] = model_stat_in_layer_class(classes_pred, labels, idx_vals, class_metrics) | |||
stats['in_struc'] = model_stat_struc(in_layers_pred, adjs_orig, layer_metrics, adjs_pos_weights, adjs_norms) | |||
stats['bet_struc'] = model_stat_struc(bet_layers_pred, bet_adjs_orig, layer_metrics, bet_pos_weights, bet_norms) | |||
for name, stat in stats.items(): | |||
dict_to_writer(stat, writer, epoch, name, 'Validation') | |||
loss_val = sum([sum(stat['loss']) for stat in stats.values()]) | |||
print('Epoch: {:04d}'.format(epoch+1), | |||
'loss_train: {:.4f}'.format(loss_train.item()), | |||
'time: {:.4f}s'.format(time.time() - t)) | |||
def test(): | |||
# Test | |||
model.eval() | |||
classes_pred, bet_layers_pred, in_layers_pred = model(features, adjs) | |||
stats = dict() | |||
stats['in_class'] = model_stat_in_layer_class(classes_pred, labels, idx_tests, class_metrics) | |||
for name, stat in stats.items(): | |||
dict_to_writer(stat, writer, epoch, name, 'Test') | |||
loss_test = sum([sum(stat['loss']) for stat in stats.values()]) | |||
print("Test set results:", | |||
"loss= {:.4f}".format(loss_test.item())) | |||
if __name__=='__main__': | |||
# Training settings | |||
parser = argparse.ArgumentParser() | |||
parser.add_argument('--no-cuda', action='store_true', default=True, | |||
help='Disables CUDA training.') | |||
parser.add_argument('--fastmode', action='store_true', default=False, | |||
help='Validate during training pass.') | |||
parser.add_argument('--seed', type=int, default=42, help='Random seed.') | |||
parser.add_argument('--epochs', type=int, default=20, | |||
help='Number of epochs to train.') | |||
parser.add_argument('--weight_decay', type=float, default=5e-4, | |||
help='Weight decay (L2 loss on parameters).') | |||
parser.add_argument('--dropout', type=float, default=0.5, | |||
help='Dropout rate (1 - keep probability).') | |||
args = parser.parse_args() | |||
args.cuda = not args.no_cuda and torch.cuda.is_available() | |||
dataset_str = "infra" | |||
# parameter | |||
# All combination of parameters will be tested | |||
# adj_weights contains a list of different configs, for example [[1,1,1]] contains one config with equal weights | |||
# for a three-layered input | |||
adj_weights = [] | |||
# parameter, weight of link reconstruction loss function | |||
wlambdas = [10] | |||
# parameter, hidden dimension for every layer should be defined | |||
hidden_structures = [[[32],[32],[32]]] | |||
# hidden_structures = [[[16], [16], [16]],[[32], [32], [32]],[[64], [64], [64]],[[128], [128], [128]],[[256], [256], [256]]] | |||
# Learning rate | |||
lrs = [0.01] | |||
# test size | |||
test_sizes = [0.2] | |||
# Load data | |||
adjs, adjs_orig, adjs_sizes, adjs_pos_weights, adjs_norms, bet_pos_weights, bet_norms, bet_adjs, bet_adjs_orig, bet_adjs_sizes, \ | |||
features, features_sizes, labels, labels_nclass = load_data(path="../data/" + dataset_str + "/", dataset=dataset_str) | |||
# Number of layers | |||
n_inputs = len(adjs) | |||
# Weights of layers | |||
adj_weights.append(n_inputs * [1]) | |||
if args.cuda: | |||
for i in range(n_inputs): | |||
labels[i] = labels[i].cuda() | |||
adjs_pos_weights[i] = adjs_pos_weights[i].cuda() | |||
adjs[i] = adjs[i].cuda() | |||
adjs_orig[i] = adjs_orig[i].cuda() | |||
features[i] = features[i].cuda() | |||
for i in range(len(bet_adjs)): | |||
if not bet_adjs[i] is None: | |||
bet_adjs[i] = bet_adjs[i].cuda() | |||
bet_adjs_orig[i] = bet_adjs_orig[i].cuda() | |||
bet_pos_weights[i] = bet_pos_weights[i].cuda() | |||
# Number of runs | |||
for run in range(10): | |||
for ts in test_sizes: | |||
idx_trains, idx_vals, idx_tests = trains_vals_tests_split(n_inputs, [s[0] for s in adjs_sizes], val_size=0.1, | |||
test_size=ts, random_state=int(run + ts*100)) | |||
if args.cuda: | |||
for i in range(n_inputs): | |||
idx_trains[i] = idx_trains[i].cuda() | |||
idx_vals[i] = idx_vals[i].cuda() | |||
idx_tests[i] = idx_tests[i].cuda() | |||
# Train model | |||
t_total = time.time() | |||
for wlambda in wlambdas: | |||
for adj_weight in adj_weights: | |||
for lr in lrs: | |||
for hidden_structure in hidden_structures: | |||
temp_weight = ['{:.2f}'.format(x) for x in adj_weight] | |||
w_str = '-'.join(temp_weight) | |||
h_str = '-'.join([','.join(map(str, temp)) for temp in hidden_structure]) | |||
writer = SummaryWriter('log/' + dataset_str + '_run-'+ str(run)+ '/' + '_lambda-' + | |||
str(wlambda) + "_adj_w-" + w_str + "_LR-" + str(lr) + | |||
'_hStruc-' + h_str + '_test_size-' + str(ts)) | |||
model = CCN(n_inputs=n_inputs, | |||
inputs_nfeat=features_sizes, | |||
inputs_nhid=hidden_structure, | |||
inputs_nclass=labels_nclass, | |||
dropout=args.dropout) | |||
optimizer = optim.Adam(model.parameters(), | |||
lr=lr, weight_decay=args.weight_decay) | |||
if args.cuda: | |||
model.cuda() | |||
for epoch in range(args.epochs): | |||
train(epoch) | |||
# Testing | |||
test() | |||
print("Optimization Finished!") | |||
print("Total time elapsed: {:.4f}s".format(time.time() - t_total)) | |||
@@ -0,0 +1,276 @@ | |||
import numpy as np | |||
import scipy.sparse as sp | |||
import torch | |||
from sklearn.metrics import accuracy_score, f1_score | |||
import random | |||
from os import listdir | |||
from os.path import isfile, join | |||
from sklearn.model_selection import train_test_split | |||
def encode_onehot(labels): | |||
classes = set(labels) | |||
classes_dict = {c: np.identity(len(classes))[i, :] for i, c in | |||
enumerate(classes)} | |||
labels_onehot = np.array(list(map(classes_dict.get, labels)), | |||
dtype=np.int32) | |||
return labels_onehot | |||
def load_bet_adj(layer_num1, layer_num2, idx_map_l1, idx_map_l2, path="../data/cora/", dataset="cora"): | |||
temp = "{}{}.bet" + str(layer_num1) + "_" + str(layer_num2) | |||
# print("bet file") | |||
# print(temp) | |||
if not isfile(temp.format(path, dataset)): | |||
return None | |||
edges_unordered = np.genfromtxt(temp.format(path, dataset), dtype=np.int32) | |||
# idx1 = np.array(np.unique(edges_unordered[:,0]), dtype=np.int32) | |||
# idx2 = np.array(np.unique(edges_unordered[:, 1]), dtype=np.int32) | |||
# idx_map_l1 = {j: i for i, j in enumerate(idx1)} | |||
# idx_map_l2 = {j: i for i, j in enumerate(idx2)} | |||
N1 = len(list(idx_map_l1)) | |||
N2 = len(list(idx_map_l2)) | |||
edges = np.array(list(map(idx_map_l1.get, edges_unordered[:,0])) + list(map(idx_map_l2.get, edges_unordered[:,1])), | |||
dtype=np.int32).reshape(edges_unordered.shape, order='F') | |||
adj = sp.coo_matrix((np.ones(edges.shape[0]), (edges[:, 0], edges[:, 1])), | |||
shape=(N1, N2), | |||
dtype=np.float32) | |||
adj_orig = sparse_mx_to_torch_sparse_tensor(adj) | |||
adj = normalize(adj) | |||
adj = sparse_mx_to_torch_sparse_tensor(adj) | |||
return adj, adj_orig | |||
def load_in_adj(layer_num, idx_map=None, path="../data/cora/", dataset="cora"): | |||
temp = "{}{}.adj" + str(layer_num) | |||
edges_unordered = np.genfromtxt(temp.format(path, dataset), | |||
dtype=np.int32) | |||
if idx_map is None: | |||
idx = np.array(np.unique(edges_unordered.flatten()), dtype=np.int32) | |||
N = len(list(idx)) | |||
idx_map = {j: i for i, j in enumerate(idx)} | |||
else: | |||
N = len(list(idx_map)) | |||
edges = np.array(list(map(idx_map.get, edges_unordered.flatten())), | |||
dtype=np.int32).reshape(edges_unordered.shape) | |||
adj = sp.coo_matrix((np.ones(edges.shape[0]), (edges[:, 0], edges[:, 1])), | |||
shape=(N, N), | |||
dtype=np.float32) | |||
# print("Edges") | |||
# print(np.count_nonzero(adj.toarray())) | |||
#build symmetric adjacency matrix | |||
adj = adj + adj.T.multiply(adj.T > adj) - adj.multiply(adj.T > adj) | |||
adj_orig = sparse_mx_to_torch_sparse_tensor(adj) | |||
adj = normalize(adj + sp.eye(adj.shape[0])) | |||
adj = sparse_mx_to_torch_sparse_tensor(adj) | |||
return adj, adj_orig | |||
def load_features_labels(layer_num, path, dataset,N=-1): | |||
print('Loading {} dataset...'.format(dataset)) | |||
temp = "{}{}.feat"+str(layer_num) | |||
idx_features_labels = np.genfromtxt(temp.format(path, dataset), | |||
dtype=np.dtype(str)) | |||
temp = idx_features_labels[:, 1:-1] | |||
if temp.size == 0: | |||
features = sp.csr_matrix(np.identity(N), dtype=np.float32) | |||
else: | |||
features = sp.csr_matrix(temp, dtype=np.float32) | |||
labels = encode_onehot(idx_features_labels[:, -1]) | |||
#build graph | |||
idx = np.array(idx_features_labels[:, 0], dtype=np.int32) | |||
idx_map = {j: i for i, j in enumerate(idx)} | |||
features = normalize(features) | |||
features = sp.csr_matrix(features) | |||
features = sparse_mx_to_torch_sparse_tensor(features) | |||
# features = torch.FloatTensor(np.array(features.todense())) | |||
labels = torch.LongTensor(np.where(labels)[1]) | |||
return features, labels, idx_map | |||
def train_val_test_split(N, val_size=0.2, test_size=0.2, random_state=1): | |||
idx_train_temp, idx_test = train_test_split(range(N), test_size=test_size, random_state=random_state) | |||
if val_size == 0: | |||
idx_train = idx_train_temp | |||
else: | |||
idx_train, idx_val = train_test_split(idx_train_temp, test_size=val_size, random_state=random_state) | |||
idx_train = torch.LongTensor(idx_train) | |||
idx_val = torch.LongTensor(idx_val) | |||
idx_test = torch.LongTensor(idx_test) | |||
return idx_train, idx_val, idx_test | |||
def trains_vals_tests_split(n_layers, labels_sizes, val_size, test_size, random_state): | |||
idx_trains = [] | |||
idx_vals = [] | |||
idx_tests = [] | |||
for i in range(n_layers): | |||
idx_train, idx_val, idx_test = train_val_test_split(labels_sizes[i], val_size, test_size, random_state) | |||
idx_trains.append(idx_train) | |||
idx_vals.append(idx_val) | |||
idx_tests.append(idx_test) | |||
return idx_trains, idx_vals, idx_tests | |||
def load_data(path="../data/cora/", dataset="cora"): | |||
layers = 0 | |||
adjs = [] | |||
adjs_orig = [] | |||
adjs_sizes = [] | |||
adjs_pos_weights = [] | |||
adjs_norms = [] | |||
bet_adjs = [] | |||
bet_adjs_orig = [] | |||
bet_adjs_sizes = [] | |||
bet_pos_weights = [] | |||
bet_norms = [] | |||
features = [] | |||
features_sizes = [] | |||
labels = [] | |||
labels_nclass = [] | |||
idx_maps = [] | |||
for f in listdir(path): | |||
if isfile(join(path, f)): | |||
if 'adj' in f: | |||
layers += 1 | |||
for i in range(layers): | |||
feature, label, idx_map = load_features_labels(i, path, dataset) | |||
adj, adj_orig = load_in_adj(i, idx_map, path, dataset) | |||
pos_weight = float(adj.shape[0] * adj.shape[1] - adj.to_dense().sum()) / adj.to_dense().sum() | |||
norm = adj.shape[0] * adj.shape[1] / float((adj.shape[0] * adj.shape[1] - adj.to_dense().sum()) * 2) | |||
idx_maps.append(idx_map) | |||
adjs.append(adj) | |||
adjs_orig.append(adj_orig) | |||
adjs_sizes.append(tuple(adj.size())) | |||
adjs_pos_weights.append(pos_weight) | |||
adjs_norms.append(norm) | |||
features.append(feature) | |||
features_sizes.append(feature.shape[1]) | |||
labels.append(label) | |||
labels_nclass.append((label.max().item() + 1)) | |||
for i in range(layers): | |||
for j in range(i+1,layers): | |||
bet_adj, bet_adj_orig = load_bet_adj(i, j, idx_maps[i], idx_maps[j], path, dataset) | |||
bet_adjs.append(bet_adj) | |||
bet_adjs_orig.append(bet_adj_orig) | |||
bet_adjs_sizes.append(tuple(bet_adj.size()) if not bet_adj is None else tuple((0,0))) | |||
if not bet_adj is None: | |||
pos_weight = float( | |||
bet_adj.shape[0] * bet_adj.shape[1] - bet_adj.to_dense().sum()) / bet_adj.to_dense().sum() | |||
norm = bet_adj.shape[0] * bet_adj.shape[1] / float((bet_adj.shape[0] * bet_adj.shape[1] - bet_adj.to_dense().sum()) * 2) | |||
else: | |||
norm = None | |||
pos_weight = None | |||
bet_pos_weights.append(pos_weight) | |||
bet_norms.append(norm) | |||
return adjs, adjs_orig, adjs_sizes, adjs_pos_weights, adjs_norms, bet_pos_weights, bet_norms, bet_adjs, bet_adjs_orig, \ | |||
bet_adjs_sizes, features, features_sizes, labels, labels_nclass | |||
def normalize(mx): | |||
"""Row-normalize sparse matrix""" | |||
rowsum = np.array(mx.sum(1)) | |||
r_inv = np.power(rowsum, -1).flatten() | |||
r_inv[np.isinf(r_inv)] = 0. | |||
r_mat_inv = sp.diags(r_inv) | |||
mx = r_mat_inv.dot(mx) | |||
return mx | |||
def class_accuracy(output, labels, type=None): | |||
preds = output.max(1)[1].type_as(labels) | |||
return accuracy_score(labels.data, preds) | |||
def class_f1(output, labels, type='micro'): | |||
preds = output.max(1)[1].type_as(labels) | |||
return f1_score(labels.data, preds, average=type) | |||
def layer_accuracy(output, real, type=None): | |||
preds = output.data.clone() | |||
true = real.data.clone() | |||
preds[output < 0.5] = 0 | |||
preds[output >= 0.5] = 1 | |||
true[real > 0] = 1 | |||
return accuracy_score(true, preds) | |||
def layer_f1(output, real, type='micro'): | |||
preds = output.data.clone() | |||
true = real.data.clone() | |||
preds[output < 0.5] = 0 | |||
preds[output >= 0.5] = 1 | |||
true[real > 0] = 1 | |||
return f1_score(true, preds, average=type) | |||
def writer_data(values, writer, epoch, type, name): | |||
try: | |||
for i, j in enumerate(values): | |||
# my_dic[name+str(i)] = j | |||
writer.add_scalar(type + "/" + name + str(i), j, epoch) | |||
except TypeError as te: | |||
writer.add_scalar(type + "/" + name, values, epoch) | |||
# type = in_class, in_struc, bet_struc | |||
def dict_to_writer(stats, writer, epoch, type, train_test_val): | |||
for key, value in stats.items(): | |||
# type_str = train_test_val + "/" + type + '/Loss' if key == 'loss' else train_test_val + type + '/Stats' | |||
type_str = train_test_val + "/" + type | |||
writer_data(value, writer, epoch, type_str, key) | |||
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 sparse_to_tuple(sparse_mx): | |||
if not sp.isspmatrix_coo(sparse_mx): | |||
sparse_mx = sparse_mx.tocoo() | |||
coords = np.vstack((sparse_mx.row, sparse_mx.col)).transpose() | |||
values = sparse_mx.data | |||
shape = sparse_mx.shape | |||
return coords, values, shape | |||
def between_index(n_inputs, i,j): | |||
return int(i * n_inputs - (i * (i + 1) / 2) + (j - i - 1)) | |||
def gather_edges(pos_edges, neg_edges): | |||
all_edges = [[], []] | |||
all_edges[0].extend([idx_i[0] for idx_i in pos_edges]) | |||
all_edges[1].extend([idx_i[1] for idx_i in pos_edges]) | |||
all_edges[0].extend([idx_i[0] for idx_i in neg_edges]) | |||
all_edges[1].extend([idx_i[1] for idx_i in neg_edges]) | |||
return all_edges | |||