author int64 658 755k | date stringlengths 19 19 | timezone int64 -46,800 43.2k | hash stringlengths 40 40 | message stringlengths 5 490 | mods list | language stringclasses 20 values | license stringclasses 3 values | repo stringlengths 5 68 | original_message stringlengths 12 491 |
|---|---|---|---|---|---|---|---|---|---|
87,709 | 20.03.2021 21:22:19 | 0 | ba47bd62273f15a5d5f242f2930dbdd2f876b08f | more GMAN layers that might be useful | [
{
"change_type": "MODIFY",
"old_path": "docs/source/modules/root.rst",
"new_path": "docs/source/modules/root.rst",
"diff": "@@ -102,4 +102,4 @@ Auxiliary Graph Convolutional Layers\n.. automodule:: torch_geometric_temporal.nn.convolutional.gman\n:members:\n:undoc-members:\n- :exclude-members: conv2d... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | more GMAN layers that might be useful |
87,709 | 20.03.2021 21:37:58 | 0 | e353f1010a4653ff3cd160aea15ddae0b82c98c3 | more tests for GMAN layers | [
{
"change_type": "MODIFY",
"old_path": "examples/GMAN_components_example.py",
"new_path": "examples/GMAN_components_example.py",
"diff": "-import torch.nn as nn\nimport torch\n-import math\nimport networkx as nx\nimport numpy as np\n-from torch_geometric_temporal.nn import STEmbedding, STAttBlock, t... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | more tests for GMAN layers |
87,709 | 20.03.2021 21:42:41 | 0 | 1bb9d5bc7d7b2206b9d3436e384f3bc1dea193d2 | README MTGNN and GMAN auxiliary layers | [
{
"change_type": "MODIFY",
"old_path": "README.md",
"new_path": "README.md",
"diff": "@@ -123,6 +123,22 @@ In detail, the following temporal graph neural networks were implemented.\n* **[DConv](https://pytorch-geometric-temporal.readthedocs.io/en/latest/modules/root.html#torch_geometric_temporal.nn.... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | README MTGNN and GMAN auxiliary layers |
87,709 | 20.03.2021 22:29:49 | 0 | 8b4a966b12ae1afaad788a97d19651d21cb7debd | delete some auxiliary layers | [
{
"change_type": "MODIFY",
"old_path": "docs/source/modules/root.rst",
"new_path": "docs/source/modules/root.rst",
"diff": "@@ -74,7 +74,7 @@ Discrete Temporal Graph Convolutional Layers\n.. automodule:: torch_geometric_temporal.nn.convolutional.gman\n:members:\n:undoc-members:\n- :exclude-members: ... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | delete some auxiliary layers |
87,721 | 22.03.2021 15:42:12 | -3,600 | 9790eb73eb29a1b2f6e477c293522ecdce84c343 | add wikimaths dataset loader | [
{
"change_type": "ADD",
"old_path": null,
"new_path": "torch_geometric_temporal/data/dataset/wikimath.py",
"diff": "+import io\n+import json\n+import numpy as np\n+from six.moves import urllib\n+from torch_geometric_temporal.data.discrete.static_graph_discrete_signal import StaticGraphDiscreteSignal... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | add wikimaths dataset loader |
87,721 | 22.03.2021 15:42:36 | -3,600 | cafeaf00f93ae6d6dcd2865e631ab213b3bcc203 | update init with wikimaths dataset | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/data/dataset/__init__.py",
"new_path": "torch_geometric_temporal/data/dataset/__init__.py",
"diff": "@@ -2,3 +2,4 @@ from .chickenpox import ChickenpoxDatasetLoader\nfrom .pedalme import PedalMeDatasetLoader\nfrom .metr_la import METRL... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | update init with wikimaths dataset |
87,721 | 22.03.2021 17:28:05 | -3,600 | 6007d5984f21c1a6104fda78603a4c97c8936621 | update loader with lagged target features | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/data/dataset/wikimath.py",
"new_path": "torch_geometric_temporal/data/dataset/wikimath.py",
"diff": "@@ -26,24 +26,25 @@ class WikiMathsDatasetLoader(object):\ndef _get_edge_weights(self):\nself._edge_weights = np.array(self._dataset[\... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | update loader with lagged target features |
87,709 | 24.03.2021 10:24:24 | 0 | 85fbc822feb0f2ef220bfdd0065c0563e76ec53c | method names update | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/convolutional/gman.py",
"new_path": "torch_geometric_temporal/nn/convolutional/gman.py",
"diff": "@@ -15,20 +15,20 @@ class conv2d_(nn.Module):\nself.padding_size = math.ceil(kernel_size)\nelse:\nself.padding_size = [0, 0]\n- self.c... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | method names update |
87,709 | 24.03.2021 23:04:43 | 0 | 0141a0536c9c1b1713be2d36955e805a5fb7a9c0 | X_in to replace input | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"new_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"diff": "@@ -380,12 +380,12 @@ class MTGNN(nn.Module):\nself._idx = torch.arange(self._num_nodes)\n- def forward(self, input: torch.FloatTens... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | X_in to replace input |
87,709 | 24.03.2021 23:05:48 | 0 | dff37950c9b34ef6da99e870587a9e054e407f15 | adj for all adjacency matrices | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"new_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"diff": "@@ -17,18 +17,18 @@ class Conv(nn.Module):\ndef __init__(self):\nsuper(Conv, self).__init__()\n- def forward(self, X: torch.FloatTen... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | adj for all adjacency matrices |
87,709 | 24.03.2021 23:10:07 | 0 | 70aaaa3f6125307229c52cfa77282ce085fbe280 | remove unused in test | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -218,8 +218,7 @@ def test_mtgnn():\nnum_split = 1\ndevice = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n- torch.set_num_threads(3)\n- x, edge_index = creat... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | remove unused in test |
87,709 | 25.03.2021 15:29:59 | 0 | a62bd81f21316f6d36c66c5692d1645bfc7935d2 | remove Conv typo | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"new_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"diff": "@@ -52,7 +52,6 @@ class MixProp(nn.Module):\ndef __init__(self, c_in: int, c_out: int, gdep: int, dropout: float, alpha: float):\nsu... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | remove Conv typo |
87,709 | 25.03.2021 15:35:57 | 0 | 2995bbe0ea41e33175133eefde748771c07e96e9 | remove unnecessary loops in test | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -221,11 +221,7 @@ def test_mtgnn():\n_, edge_index = create_mock_data(number_of_nodes=num_nodes, edge_per_node=8, in_channels=in_dim)\nmock_adj = to_scipy_sparse_matrix(edge_ind... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | remove unnecessary loops in test |
87,709 | 25.03.2021 15:53:40 | 0 | 16b577169c86c8128c8ca5f90d7d5141434b4438 | MixProp update | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -197,7 +197,7 @@ def test_mtgnn():\nTesting MTGNN block\n\"\"\"\ngcn_true = True\n- buildA_true = True\n+ build_adj = True\ndropout = 0.3\nsubgraph_size = 20\ngcn_depth = 2\n@@ ... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | MixProp update |
87,709 | 25.03.2021 15:58:39 | 0 | 2c93580d9994faed8d91a2a4b25fa046a1d2bfb8 | kernel set hyper | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -216,16 +216,16 @@ def test_mtgnn():\npropalpha = 0.05\ntanhalpha = 3\nnum_split = 1\n+ kernel_set = [2, 3, 6, 7]\ndevice = torch.device(\"cuda\" if torch.cuda.is_available() el... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | kernel set hyper |
87,709 | 25.03.2021 16:11:30 | 0 | b3b2e22a83dde0ac4d2b189c00cbd694af6b125c | data generation in test: pure pytorch | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -2,7 +2,7 @@ import torch\nimport numpy as np\nimport networkx as nx\nfrom torch_geometric.data import Data\n-from torch_geometric.utils import to_scipy_sparse_matrix\n+from tor... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | data generation in test: pure pytorch |
87,709 | 25.03.2021 16:17:44 | 0 | 053033b0be36dde124edb9e6d68af7c1d688246d | Captital data matrices | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"new_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"diff": "@@ -184,23 +184,23 @@ class GraphConstructor(nn.Module):\nelse:\nnn.init.uniform_(p)\n- def forward(self, idx: torch.LongTensor, sta... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | Captital data matrices |
87,709 | 25.03.2021 16:22:32 | 0 | 2d3a3c34297dc9e7fa269e81d554a5760c528f18 | kernel size hyper | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -217,13 +217,14 @@ def test_mtgnn():\ntanhalpha = 3\nnum_split = 1\nnum_edges = 10\n+ kernel_size = 7\nkernel_set = [2, 3, 6, 7]\ndevice = torch.device(\"cuda\" if torch.cuda.is... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | kernel size hyper |
87,709 | 25.03.2021 16:26:12 | 0 | f56739f529e99c2321f9e83567f8f69cc72aa78a | kernel size hyper, line break | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"new_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"diff": "@@ -281,6 +281,7 @@ class MTGNN(nn.Module):\ngcn_depth (int) : Graph convolution depth.\nnum_nodes (int) : Number of nodes in the gr... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | kernel size hyper, line break |
87,709 | 25.03.2021 16:49:01 | 0 | c454928ff41892251c56b94d13c94d39420cda7f | Capital matrices, line break | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"new_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"diff": "@@ -431,29 +431,27 @@ class MTGNN(nn.Module):\nif self._gcn_true:\nif self._build_adj_true:\nif idx is None:\n- adp = self._graph_co... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | Capital matrices, line break |
87,709 | 25.03.2021 16:54:09 | 0 | 53f0df21bbe689e35fe15115ba1d72629b71d13e | remove a loop | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"new_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"diff": "@@ -329,12 +329,7 @@ class MTGNN(nn.Module):\nelse:\nself._receptive_field = layers*(kernel_size-1) + 1\n- for i in range(1):\n- if ... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | remove a loop |
87,709 | 25.03.2021 17:38:57 | 0 | 069f62c49fadba1b38004ab4d1b7388b87c4d916 | one more mtgnn layer | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"new_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"diff": "@@ -270,6 +270,91 @@ class LayerNormalization(nn.Module):\n'elementwise_affine={elementwise_affine}'.format(**self.__dict__)\n+class... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | one more mtgnn layer |
87,709 | 25.03.2021 17:50:45 | 0 | 2874c488419b8f3798d8fd7615aebdff1663977c | MTGNNLayer docstrings | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"new_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"diff": "@@ -276,10 +276,24 @@ class MTGNNLayer(nn.Module):\n<https://arxiv.org/pdf/2005.11650.pdf>`_\nArgs:\n- c_in (int): Number of input c... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | MTGNNLayer docstrings |
87,709 | 25.03.2021 21:49:04 | 0 | ccfb841dce4a686b20b6d7891e63ff782e44def4 | fix warning from astgcn | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/convolutional/astgcn.py",
"new_path": "torch_geometric_temporal/nn/convolutional/astgcn.py",
"diff": "@@ -67,7 +67,7 @@ class ChebConvAttention(MessagePassing):\ndef _reset_parameters(self):\nnn.init.xavier_uniform_(self.weight)\n- ... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | fix warning from astgcn |
87,709 | 25.03.2021 23:25:26 | 0 | 2cd0a4e8537e772733e0f52575c977604c95a7c2 | MTGNN docstrings | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"new_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"diff": "@@ -184,23 +184,23 @@ class GraphConstructor(nn.Module):\nelse:\nnn.init.uniform_(p)\n- def forward(self, idx: torch.LongTensor, FE:... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | MTGNN docstrings |
87,709 | 26.03.2021 09:46:39 | 0 | 2532eb3778fb472ecdd0b9bf8e09082b04656248 | MTGNN coverage 92% to 94% | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -230,6 +230,15 @@ def test_mtgnn():\nskip_channels=skip_channels, end_channels=end_channels,\nseq_length=seq_in_len, in_dim=in_dim, out_dim=seq_out_len,\nlayers=layers, propalph... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | MTGNN coverage 92% to 94% |
87,709 | 26.03.2021 10:00:18 | 0 | 4fad62a09c7acea51333deea899dd93ba186f3ce | mtgnn input fix in test | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -221,7 +221,8 @@ def test_mtgnn():\nkernel_set = [2, 3, 6, 7]\ndevice = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n- Tilde_A = barabasi_albert_graph(num_n... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | mtgnn input fix in test |
87,709 | 26.03.2021 10:02:54 | 0 | f8597f3f9c45aec5fb9a83215eb649ce74c1ed5e | mtgnn coverage 94%->95% | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -264,7 +264,7 @@ def test_mtgnn():\noutput_with_feat = model_with_feat(tx, A_tilde, idx=id, FE=FE)\noutput_with_feat = output_with_feat.transpose(1, 3)\nassert output_with_feat.... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | mtgnn coverage 94%->95% |
87,709 | 26.03.2021 10:12:42 | 0 | 50f22650a8c6c1d55a8fa6cdd779e47278258581 | MTGNN coverage 95%->97% | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -268,6 +268,27 @@ def test_mtgnn():\noutput3 = output3.transpose(1, 3)\nassert output3.shape == (batch_size, 1, num_nodes, seq_out_len)\n+ seq_in_len = 24\n+ x_all = 2 * torch.r... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | MTGNN coverage 95%->97% |
87,709 | 26.03.2021 10:21:29 | 0 | 6013095231aecc198bfe2725ac5c989fc4f35b16 | MTGNN coverage 97%->98% | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -233,7 +233,7 @@ def test_mtgnn():\nlayers=layers, propalpha=propalpha, tanhalpha=tanhalpha, layer_norm_affline=True)\nxd = 8\nFE = torch.rand(num_nodes, xd)\n- model_with_feat ... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | MTGNN coverage 97%->98% |
87,709 | 26.03.2021 10:24:59 | 0 | 15a542f6404e541ab8e2a33c2264b6d8d29262e8 | MTGNN coverage 98%->99% | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -287,6 +287,13 @@ def test_mtgnn():\nskip_channels=skip_channels, end_channels=end_channels,\nseq_length=seq_in_len, in_dim=in_dim, out_dim=seq_out_len,\nlayers=layers, propalph... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | MTGNN coverage 98%->99% |
87,709 | 26.03.2021 10:28:42 | 0 | e37ef1f4a2c3e1579153156561356602531fc149 | MTGNN test typo | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -248,6 +248,13 @@ def test_mtgnn():\nskip_channels=skip_channels, end_channels=end_channels,\nseq_length=seq_in_len, in_dim=in_dim, out_dim=seq_out_len,\nlayers=layers, propalph... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | MTGNN test typo |
87,709 | 26.03.2021 10:39:31 | 0 | 547c1ca65349d9b9e3ad571cc001830fac7ec0a9 | MTGNN coverage 99%, 3 missing cases left | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -250,7 +250,7 @@ def test_mtgnn():\nlayers=layers, propalpha=propalpha, tanhalpha=tanhalpha, layer_norm_affline=True, xd=xd)\nmodel4 = MTGNN(gcn_true=False, build_adj=build_adj,... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | MTGNN coverage 99%, 3 missing cases left |
87,709 | 26.03.2021 11:06:13 | 0 | 863aafc5286dbd186d0ab3dd3c9925f282275fa5 | hyperparameters added to gman | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -337,13 +337,18 @@ def test_gman():\nbatch_size = 32\nbn_decay = 0.1\nsteps_per_day = 288\n+ same_padding = False\n+ use_bias = True\ntrainX = torch.rand(num_sample,num_his, num... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | hyperparameters added to gman |
87,709 | 26.03.2021 11:30:09 | 0 | ac68dc9c521d3cb90415fe911c6b6a9282958f03 | improve GMAN coverage by 1 case and remove unused variable | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -337,13 +337,12 @@ def test_gman():\nbatch_size = 32\nbn_decay = 0.1\nsteps_per_day = 288\n- same_padding = False\nuse_bias = True\ntrainX = torch.rand(num_sample,num_his, num_n... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | improve GMAN coverage by 1 case and remove unused variable |
87,709 | 26.03.2021 11:56:07 | 0 | cb8f292319e18a514da812840d281328c40894f7 | GMAN coverage 94%->99%, 3 missing cases | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -192,6 +192,36 @@ def test_mstgcn():\nassert outputs1.shape == (batch_size, node_count, num_for_predict)\nassert outputs2.shape == (batch_size, node_count, num_for_predict)\n+de... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | GMAN coverage 94%->99%, 3 missing cases |
87,709 | 26.03.2021 11:58:45 | 0 | f2252c218dadad993a9304d0cc49f132b0f5e35e | GMAN coverage 100%! | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/convolutional/gman.py",
"new_path": "torch_geometric_temporal/nn/convolutional/gman.py",
"diff": "@@ -59,14 +59,14 @@ class FullyConnected(nn.Module):\n<https://arxiv.org/pdf/1911.08415.pdf>`_\nArgs:\n- input_dims (int, list or tupl... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | GMAN coverage 100%! |
87,709 | 26.03.2021 12:07:29 | 0 | 4a3fe84d9e30479170d11e44a8a69d2e683e370f | astgcn variable names | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/convolutional/astgcn.py",
"new_path": "torch_geometric_temporal/nn/convolutional/astgcn.py",
"diff": "@@ -53,21 +53,22 @@ class ChebConvAttention(MessagePassing):\nassert K > 0\nassert normalization in [None, 'sym', 'rw'], 'Invalid ... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | astgcn variable names |
87,709 | 26.03.2021 12:41:37 | 0 | 262db2b4b0b6877dda3378563cd4b5fd15a47c48 | astgcn normalization | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -116,9 +116,16 @@ def test_astgcn():\nnb_chev_filter = 64\nnb_time_filter = 64\nbatch_size = 32\n+ normalization = None\n+ bias = True\nx, edge_index = create_mock_data(node_cou... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | astgcn normalization |
87,709 | 26.03.2021 13:05:06 | 0 | 8b1490d5dcc9cfa91cbb7f7d44990f7d9e117984 | ChebConvAttention test | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -4,7 +4,7 @@ import networkx as nx\nfrom torch_geometric.data import Data\nfrom torch_geometric.utils import barabasi_albert_graph\nfrom torch_geometric.transforms import Laplac... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | ChebConvAttention test |
87,712 | 28.03.2021 19:27:44 | -3,600 | 22099d1669248e5067f1fbc2adee30655393cd51 | [Test] Pragma no cover on download outside PTG | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/dataset/metr_la.py",
"new_path": "torch_geometric_temporal/dataset/metr_la.py",
"diff": "@@ -14,14 +14,12 @@ class METRLADatasetLoader(object):\nCounty in aggregated 5 minute intervals for 4 months between March 2012\nto June 2012.\n\"... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | [Test] Pragma no cover on download outside PTG |
87,712 | 28.03.2021 19:28:56 | -3,600 | 9b12f60b718480dd9129dbd6f09f18de6b229e60 | [Test] Pragma no cover on download outside PTG and removing __name__ == main. Coverage -> 100% | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/dataset/pems_bay.py",
"new_path": "torch_geometric_temporal/dataset/pems_bay.py",
"diff": "@@ -15,7 +15,7 @@ class PemsBayDatasetLoader(object):\nself.raw_data_dir = raw_data_dir\nself._read_web_data()\n- def _download_url(self, url, s... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | [Test] Pragma no cover on download outside PTG and removing __name__ == main. Coverage -> 100% |
87,712 | 28.03.2021 19:40:12 | -3,600 | 672b3de6fda053fc7bd2c37a0268927b9e731039 | [Docs] Paper details | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/dataset/pems_bay.py",
"new_path": "torch_geometric_temporal/dataset/pems_bay.py",
"diff": "@@ -7,7 +7,10 @@ from six.moves import urllib\nfrom ..signal import StaticGraphTemporalSignal\nclass PemsBayDatasetLoader(object):\n- \"\"\"A tr... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | [Docs] Paper details |
87,712 | 28.03.2021 19:43:45 | -3,600 | 53707547ae63ff5585571dc001356f988b9fec73 | [Docs] Dataset and Paper details | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/dataset/metr_la.py",
"new_path": "torch_geometric_temporal/dataset/metr_la.py",
"diff": "@@ -13,6 +13,10 @@ class METRLADatasetLoader(object):\nreadings collected from 207 loop detectors on highways in Los Angeles\nCounty in aggregated... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | [Docs] Dataset and Paper details |
87,712 | 28.03.2021 19:47:38 | -3,600 | 730a17815589f39aa187c342804bb391bde8e6b6 | [Docs] Dynamic network details | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/dataset/pems_bay.py",
"new_path": "torch_geometric_temporal/dataset/pems_bay.py",
"diff": "@@ -9,6 +9,11 @@ from ..signal import StaticGraphTemporalSignal\nclass PemsBayDatasetLoader(object):\n\"\"\"A traffic forecasting dataset as des... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | [Docs] Dynamic network details |
87,712 | 28.03.2021 19:48:54 | -3,600 | 37090d05fd44ab9229f30954c444d07e02c5d46d | Refactoring to match behaviour of new methods | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/recurrent/dcrnn.py",
"new_path": "torch_geometric_temporal/nn/recurrent/dcrnn.py",
"diff": "@@ -38,10 +38,6 @@ class DConv(MessagePassing):\ndef message(self, x_j, norm):\nreturn norm.view(-1, 1) * x_j\n- def __repr__(self): # pragm... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | Refactoring to match behaviour of new methods |
87,712 | 28.03.2021 19:54:52 | -3,600 | c3ba64ea8e6216f6549dfaec1aa0befb0ab0a832 | [Test] DCRNN -> 100% | [
{
"change_type": "MODIFY",
"old_path": "test/recurrent_test.py",
"new_path": "test/recurrent_test.py",
"diff": "@@ -189,6 +189,20 @@ def test_dcrnn_layer():\nassert H.shape == (number_of_nodes, out_channels)\n+ layer = DCRNN(in_channels=in_channels, out_channels=out_channels, K=3)\n+\n+ H = layer(X,... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | [Test] DCRNN -> 100% |
87,711 | 29.03.2021 23:19:14 | -3,600 | a7eb5b6c4f77de384f6b1da0102abf0acbc17988 | Added MPNN_LSTM | [
{
"change_type": "ADD",
"old_path": null,
"new_path": "torch_geometric_temporal/nn/convolutional/mpnn_lstm.py",
"diff": "+\n+import torch\n+import torch.nn as nn\n+import torch.nn.functional as F\n+from torch_geometric.nn import GCNConv\n+import networkx as nx\n+import numpy as np\n+#import scipy.sp... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | Added MPNN_LSTM |
87,711 | 30.03.2021 01:25:41 | -7,200 | d6b335edc411b26ceecf1a75c5e25148b4d9f836 | Removed output layer# to return only the representations | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/convolutional/mpnn_lstm.py",
"new_path": "torch_geometric_temporal/nn/convolutional/mpnn_lstm.py",
"diff": "@@ -22,7 +22,7 @@ class MPNN_LSTM(nn.Module):\ndropout (float): Dropout rate.\n\"\"\"\ndef __init__(self, in_channels: int, ... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | Removed output layer# to return only the representations |
87,711 | 30.03.2021 12:04:21 | -3,600 | 8bb8a40d6e81a32d3838994dd6b5a8a7319fd679 | Added loader for English Covid19 dataset | [
{
"change_type": "ADD",
"old_path": null,
"new_path": "torch_geometric_temporal/dataset/encovid.py",
"diff": "+\n+import numpy as np\n+from datetime import date, timedelta\n+import pandas as pd\n+import networkx as nx\n+from torch_geometric_temporal.signal import DynamicGraphTemporalSignal\n+from sc... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | Added loader for English Covid19 dataset |
87,711 | 30.03.2021 23:40:59 | -7,200 | 4b32c8c7c1919506d91f9d777f26983659485be5 | Fixed _init_ super | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/recurrent/mpnn_lstm.py",
"new_path": "torch_geometric_temporal/nn/recurrent/mpnn_lstm.py",
"diff": "@@ -5,8 +5,7 @@ from torch_geometric.nn import GCNConv\nclass MPNNLSTM(nn.Module):\nr\"\"\"An implementation of the Message Passing ... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | Fixed _init_ super |
87,711 | 31.03.2021 00:54:06 | -7,200 | ae689a035f5c10532b4f0c48e1edfd443bd0cc02 | Changed defined dims of output representations | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/recurrent/mpnn_lstm.py",
"new_path": "torch_geometric_temporal/nn/recurrent/mpnn_lstm.py",
"diff": "@@ -64,7 +64,7 @@ class MPNNLSTM(nn.Module):\n* **edge_weight** *(PyTorch LongTensor, optional)* - Edge weight vector.\nReturn types... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | Changed defined dims of output representations |
87,711 | 31.03.2021 00:56:31 | -7,200 | 7867e594bafbe4456ee5b14d7e27e53fc70b429e | Adjusted MPNN LSTM test | [
{
"change_type": "MODIFY",
"old_path": "test/recurrent_test.py",
"new_path": "test/recurrent_test.py",
"diff": "@@ -124,12 +124,13 @@ def test_mpnn_lstm_layer():\nlayer = MPNNLSTM(in_channels=in_channels,\nhidden_size=hidden_size,\nout_channels=out_channels,\n- num_nodes=100,\n+ num_nodes=number_of_... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | Adjusted MPNN LSTM test |
87,711 | 31.03.2021 01:23:47 | -7,200 | 9feb7fa3a97b1a87c3bf851051727ed1c3f23813 | Updated hidden representation dim definition | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/recurrent/mpnn_lstm.py",
"new_path": "torch_geometric_temporal/nn/recurrent/mpnn_lstm.py",
"diff": "@@ -64,7 +64,7 @@ class MPNNLSTM(nn.Module):\n* **edge_weight** *(PyTorch LongTensor, optional)* - Edge weight vector.\nReturn types... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | Updated hidden representation dim definition |
87,709 | 31.03.2021 12:53:41 | -3,600 | ab05e1f67be21ad5d2ff5f0ab775e329d9500954 | bibentry names | [
{
"change_type": "MODIFY",
"old_path": "README.md",
"new_path": "README.md",
"diff": "@@ -36,7 +36,7 @@ If you find *PyTorch Geometric Temporal* and the new datasets useful in your res\n```bibtex\n@misc{pytorch_geometric_temporal,\n- author = {Benedek, Rozemberczki and Paul, Scherer and Yixuan, He a... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | bibentry names |
87,709 | 31.03.2021 12:54:54 | -3,600 | 212875fac2f34687cb17f3c3a8de445a17cb114a | fix bibentry names | [
{
"change_type": "MODIFY",
"old_path": "docs/source/notes/introduction.rst",
"new_path": "docs/source/notes/introduction.rst",
"diff": "@@ -11,7 +11,7 @@ If you find *PyTorch Geometric Temporal* useful in your research, please conside\n.. code-block:: latex\n>@misc{pytorch_geometric_temporal,\n- aut... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | fix bibentry names |
87,709 | 07.04.2021 15:50:05 | -3,600 | 2e3590f69744dc1f8fe77cb87782b940fb6850f7 | reduce memory by removing unused i in for loop | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"new_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"diff": "@@ -91,10 +91,9 @@ class MixProp(nn.Module):\nH = X\nH_0 = X\nA = A / d.view(-1, 1)\n- for i in range(self._gdep):\n+ for _ in range... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | reduce memory by removing unused i in for loop |
87,709 | 07.04.2021 16:08:59 | -3,600 | 9921ac2f323cbdbf11524f25cb61c92bf1f55a1f | mtgnn cover one more line | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -324,7 +324,7 @@ def test_mtgnn():\noutput = model(tx, A_tilde, idx=id)\noutput = output.transpose(1, 3)\nassert output.shape == (batch_size, 1, num_nodes, seq_out_len)\n- outpu... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | mtgnn cover one more line |
87,709 | 07.04.2021 16:14:12 | -3,600 | b76cd229b95f5fb9691a0a06688538baf7d95873 | mtgnn one more line | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"new_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"diff": "@@ -267,10 +267,6 @@ class LayerNormalization(nn.Module):\nelse:\nreturn F.layer_norm(X, tuple(X.shape[1:]), self._weight, self._bia... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | mtgnn one more line |
87,709 | 07.04.2021 16:18:25 | -3,600 | 96c3e55c954abbf2750c79d2ad28cd70c0660473 | mtgnn coverage 100% | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"new_path": "torch_geometric_temporal/nn/convolutional/mtgnn.py",
"diff": "@@ -231,8 +231,6 @@ class LayerNormalization(nn.Module):\n\"\"\"\ndef __init__(self, normalized_shape: int, eps: float=1e-5, element... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | mtgnn coverage 100% |
87,709 | 07.04.2021 16:24:46 | -3,600 | fa9dd2239c777ae7dae7e983862d13a98c692f3e | astgcn coverage improve 1 line | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -110,6 +110,7 @@ def test_astgcn():\nattention = torch.nn.functional.softmax(torch.rand((batch_size, num_nodes, num_nodes)), dim=1)\nconv = ChebConvAttention(in_channels, out_ch... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | astgcn coverage improve 1 line |
87,709 | 07.04.2021 16:28:08 | -3,600 | 3a6b8918c28c10feb0fdd5362f0cb50471d8c180 | astgcn one more coverage | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -162,12 +162,16 @@ def test_astgcn():\nencoder_inputs, _ = batch_data\noutputs0 = model(encoder_inputs, edge_index_seq)\noutputs1 = model(encoder_inputs, edge_index_seq[0])\n- o... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | astgcn one more coverage |
87,709 | 07.04.2021 16:36:57 | -3,600 | a93e76f43926252572d84ce16de66a392eef635b | astgcn coverage also 100% | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -118,6 +118,19 @@ def test_astgcn():\nout3 = conv(x, edge_index, attention, edge_weight, lambda_max=3.0)\nassert out3.size() == (batch_size, num_nodes, out_channels)\n+ batch = ... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | astgcn coverage also 100% |
87,709 | 09.04.2021 16:00:00 | 0 | 5570ebabc9577504909439195140919dd15949a0 | astgcn and mstgcn device fixed | [
{
"change_type": "ADD",
"old_path": null,
"new_path": ".vscode/settings.json",
"diff": "+{\n+ \"python.pythonPath\": \"/home/songchaow/.conda/envs/PGT/bin/python\"\n+}\n\\ No newline at end of file\n"
},
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path":... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | astgcn and mstgcn device fixed |
87,709 | 09.04.2021 16:10:36 | 0 | a7eebfc2cfea94ab0b852d75e01b02d9102b435f | gman device fixed | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -259,8 +259,8 @@ def test_gman():\nmodel = GMAN(L, K, d, num_his, bn_decay=bn_decay, steps_per_day=steps_per_day, use_bias=use_bias, mask=mask).to(device)\nmodel2 = GMAN(L, K, d... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | gman device fixed |
87,709 | 09.04.2021 16:32:02 | 0 | adc9eb18c8283ed7157d718b3417eb4d7bd40576 | mtgnn device fixed | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -296,8 +296,7 @@ def test_mtgnn():\ndevice = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\nedge_index = barabasi_albert_graph(num_nodes, num_edges).to(device... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | mtgnn device fixed |
87,709 | 09.04.2021 16:49:03 | 0 | 595a0d30d1d7d6dfc3f953c236249fca3a2cd023 | STGCN device | [
{
"change_type": "MODIFY",
"old_path": "test/convolutional_test.py",
"new_path": "test/convolutional_test.py",
"diff": "@@ -71,12 +71,13 @@ def test_temporalconv():\nin_channels = 100\nedge_per_node = 15\nout_channels = 10\n- batch, batch_targets, edge_index, edge_weight = create_mock_batch(batch_si... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | STGCN device |
87,709 | 09.04.2021 16:55:18 | 0 | b0152b50a4a388d152c3be4afbe06c61a1e6011b | recurrent test device | [
{
"change_type": "MODIFY",
"old_path": "test/recurrent_test.py",
"new_path": "test/recurrent_test.py",
"diff": "@@ -57,10 +57,13 @@ def test_gconv_lstm_layer():\nout_channels = 16\nK = 2\n+ device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\nX, edge_index = create_mock_data(nu... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | recurrent test device |
87,709 | 09.04.2021 17:05:44 | 0 | e8eaab6990e5779d4e0be21f58ceba5e0cec0f9a | more device tests on recurrent | [
{
"change_type": "MODIFY",
"old_path": "test/recurrent_test.py",
"new_path": "test/recurrent_test.py",
"diff": "@@ -151,10 +151,13 @@ def test_tgcn_layer():\nin_channels = 64\nout_channels = 16\n+ device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\nX, edge_index = create_mock_... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | more device tests on recurrent |
87,709 | 09.04.2021 17:07:35 | 0 | 4c6488bb2e82058ee21abb2d26fc9c7b45c6a9cf | recurrent tests | [
{
"change_type": "MODIFY",
"old_path": "test/recurrent_test.py",
"new_path": "test/recurrent_test.py",
"diff": "@@ -327,14 +327,17 @@ def test_dygrencoder_layer():\nlstm_out_channels = 8\nlstm_num_layers = 1\n+ device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\nX, edge_index ... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | recurrent tests |
87,709 | 09.04.2021 17:17:40 | 0 | a8f72584c7da4386a01442f5d56115c60ea8f560 | dcrnn test for device | [
{
"change_type": "MODIFY",
"old_path": "test/recurrent_test.py",
"new_path": "test/recurrent_test.py",
"diff": "@@ -212,11 +212,13 @@ def test_dcrnn_layer():\nin_channels = 64\nout_channels = 16\nK = 2\n-\n+ device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\nX, edge_index = c... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | dcrnn test for device |
87,710 | 20.04.2021 19:07:29 | -3,600 | 2cd4985e6bab86af77488616800fc917edb73e90 | Arxiv linking | [
{
"change_type": "MODIFY",
"old_path": "README.md",
"new_path": "README.md",
"diff": "If you find *PyTorch Geometric Temporal* and the new datasets useful in your research, please consider adding the following citation:\n```bibtex\n-@misc{pytorch_geometric_temporal,\n- author = {Benedek, Rozemberczk... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | Arxiv linking |
87,712 | 07.05.2021 20:23:59 | -3,600 | 3cf86f3555165474a494a056bd4459814b7c7e54 | STGCN PQ hadamard product correction | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/attention/stgcn.py",
"new_path": "torch_geometric_temporal/nn/attention/stgcn.py",
"diff": "@@ -36,7 +36,7 @@ class TemporalConv(nn.Module):\nX = X.permute(0, 3, 2, 1)\nP = self.conv_1(X)\nQ = torch.sigmoid(self.conv_2(X))\n- PQ = P... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | STGCN PQ hadamard product correction |
87,712 | 07.05.2021 22:42:14 | -3,600 | 456e55c0d297c09fa568b0190dc298ba2c44390b | DCConv reverse diffusion with in-norm | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/recurrent/dcrnn.py",
"new_path": "torch_geometric_temporal/nn/recurrent/dcrnn.py",
"diff": "import math\nimport torch\n-from torch_geometric.utils import to_dense_adj\n+from torch_geometric.utils import to_dense_adj, dense_to_sparse... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | DCConv reverse diffusion with in-norm |
87,716 | 05.06.2021 18:04:42 | -7,200 | b872b110c3d8d89e8e5e999efc56a481ca0add27 | added 2s-AGCN implementation | [
{
"change_type": "ADD",
"old_path": null,
"new_path": "torch_geometric_temporal/nn/attention/2sagcn.py",
"diff": "+import math\n+\n+import numpy as np\n+import torch\n+import torch.nn as nn\n+from torch.autograd import Variable\n+\n+\n+def edge2mat(link, num_node):\n+ A = np.zeros((num_node, num_nod... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | added 2s-AGCN implementation |
87,716 | 06.06.2021 13:03:55 | -7,200 | ff8af677a69d2d5145c37d04525d8c807f4bc7b4 | added 2s-agcn docs, test, reworked classes and functions, numpy functions to torch | [
{
"change_type": "MODIFY",
"old_path": "README.md",
"new_path": "README.md",
"diff": "@@ -120,6 +120,8 @@ In detail, the following temporal graph neural networks were implemented.\n* **[MTGNN](https://pytorch-geometric-temporal.readthedocs.io/en/latest/modules/root.html#torch_geometric_temporal.nn.a... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | added 2s-agcn docs, test, reworked classes and functions, numpy functions to torch |
87,716 | 09.06.2021 09:50:22 | -7,200 | d0100b7f28a258abd39929c38d58fbea8841e3b6 | removed mtm1 name_main | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/dataset/mtm.py",
"new_path": "torch_geometric_temporal/dataset/mtm.py",
"diff": "@@ -79,15 +79,3 @@ class MTMDatasetLoader():\nself._edges, self._edge_weights, self.features, self.targets\n)\nreturn dataset\n-\n-if __name__ == \"__main... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | removed mtm1 name_main |
87,708 | 10.06.2021 22:10:42 | 10,800 | 8ff57d110b55fe5e4f8c09c23dc8bbcb98b6ad12 | Requiered Python version to installation docs | [
{
"change_type": "MODIFY",
"old_path": "README.md",
"new_path": "README.md",
"diff": "@@ -143,6 +143,7 @@ If you notice anything unexpected, please open an [issue](https://benedekrozembe\n**Installation**\n+Binaries are provided for Python version <= 3.8.\n**PyTorch 1.8.0**\n"
},
{
"change_t... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | Requiered Python version to installation docs |
87,716 | 11.06.2021 09:15:13 | -7,200 | a3bdcbbd6b9c0c63e94c3f234119ed2f7cea09ce | fixed links, fixed tutorial | [
{
"change_type": "MODIFY",
"old_path": "docs/source/notes/introduction.rst",
"new_path": "docs/source/notes/introduction.rst",
"diff": "@@ -108,13 +108,13 @@ We provide functions to create temporal splits of the data iterators. These func\n.. code-block:: python\nfrom torch_geometric_temporal.datase... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | fixed links, fixed tutorial |
87,716 | 11.06.2021 09:39:04 | -7,200 | 7b1d53891423d7ed5474c644108aff2fc74ddefd | code coverage tsagcn | [
{
"change_type": "MODIFY",
"old_path": "test/attention_test.py",
"new_path": "test/attention_test.py",
"diff": "@@ -397,7 +397,7 @@ def test_mtgnn():\noutput3 = output3.transpose(1, 3)\nassert output3.shape == (batch_size, 1, num_nodes, seq_out_len)\n-def test_aagcn():\n+def test_tsagcn():\n\"\"\"\n... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | code coverage tsagcn |
87,709 | 16.06.2021 10:26:13 | -3,600 | 72af76073b428db951b7190bf7b667658c585e54 | GMAN test multiple blocks | [
{
"change_type": "MODIFY",
"old_path": "test/attention_test.py",
"new_path": "test/attention_test.py",
"diff": "@@ -260,6 +260,7 @@ def test_gman():\nSE = SE.to(device)\ntrainTE = (2 * torch.rand((num_sample, num_his + num_pred, 2)) - 1).to(device)\nmodel = GMAN(L, K, d, num_his, bn_decay=bn_decay, ... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | GMAN test multiple blocks |
87,710 | 13.07.2021 21:26:32 | -3,600 | 2be50982885458c2a24caee832bf27629ca7f568 | Delete .travis.yml.save | [
{
"change_type": "DELETE",
"old_path": ".travis.yml.save",
"new_path": null,
"diff": "-version: 2\n-\n-build:\n- image: latest\n-\n-python:\n- version: 3.7\n- system_packages: true\n- install:\n- - requirements: docs/requirements.txt\n- - method: setuptools\n- path: .\n-\n-formats: []\n"
}
] | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | Delete .travis.yml.save |
87,709 | 02.08.2021 14:21:51 | 0 | 47122f193c1cd7f6ed823f3e54baddaae33ef647 | GMAN T not hardcoding | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/attention/gman.py",
"new_path": "torch_geometric_temporal/nn/attention/gman.py",
"diff": "@@ -137,7 +137,7 @@ class SpatioTemporalEmbedding(nn.Module):\nfor i in range(TE.shape[0]):\ndayofweek[i] = F.one_hot(TE[..., 0][i].to(torch.i... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | GMAN T not hardcoding |
87,709 | 30.08.2021 12:15:00 | -3,600 | 789ac59a76d6478d8faf553662a6e362b4959d01 | torch 1.9 install | [
{
"change_type": "MODIFY",
"old_path": "README.md",
"new_path": "README.md",
"diff": "@@ -174,6 +174,7 @@ pip install torch-sparse -f https://pytorch-geometric.com/whl/torch-1.9.0+${CUDA\npip install torch-cluster -f https://pytorch-geometric.com/whl/torch-1.9.0+${CUDA}.html\npip install torch-splin... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | torch 1.9 install |
87,709 | 03.09.2021 15:49:17 | -3,600 | 1e21a01c8f09324950b474ad33e53714db412a38 | enable updates of recurrent weight for conv weights | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/recurrent/evolvegcno.py",
"new_path": "torch_geometric_temporal/nn/recurrent/evolvegcno.py",
"diff": "+import math\n+from typing import Optional, Tuple\n+\nimport torch\n-from torch.nn import LSTM\n-from torch_geometric.nn import GC... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | enable updates of recurrent weight for conv weights |
87,709 | 03.09.2021 15:49:42 | -3,600 | 840546a7ceda95fef3a3c75309cfed5d9a6ba5f9 | enable updates of recurrent weights for conv weights, H version | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/recurrent/evolvegcnh.py",
"new_path": "torch_geometric_temporal/nn/recurrent/evolvegcnh.py",
"diff": "@@ -3,6 +3,8 @@ from torch.nn import GRU\nfrom torch_geometric.nn import GCNConv\nfrom torch_geometric.nn import TopKPooling\n+fro... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | enable updates of recurrent weights for conv weights, H version |
87,709 | 03.09.2021 15:49:58 | -3,600 | 3bf4c103d83b119ec5454253da4142821f15f5b7 | update O version example for evolvegcn | [
{
"change_type": "MODIFY",
"old_path": "examples/recurrent/evolvegcno_example.py",
"new_path": "examples/recurrent/evolvegcno_example.py",
"diff": "@@ -26,6 +26,8 @@ class RecurrentGCN(torch.nn.Module):\nreturn h\nmodel = RecurrentGCN(node_features = 4)\n+for param in model.parameters():\n+ param.re... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | update O version example for evolvegcn |
87,709 | 06.09.2021 11:26:01 | -3,600 | 4d736ac06d0e0ebda4c3b39989a9766899c74d9f | typo of the module name | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/recurrent/evolvegcnh.py",
"new_path": "torch_geometric_temporal/nn/recurrent/evolvegcnh.py",
"diff": "import torch\nfrom torch.nn import GRU\n-from torch_geometric.nn import GCNConv\nfrom torch_geometric.nn import TopKPooling\nfrom ... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | typo of the module name |
87,709 | 06.09.2021 11:30:47 | -3,600 | 7a864081356933cbf6e4050d9c35b81349696c80 | resolve import issue | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/recurrent/evolvegcnh.py",
"new_path": "torch_geometric_temporal/nn/recurrent/evolvegcnh.py",
"diff": "+import math\n+from typing import Optional, Tuple\n+\nimport torch\n+from torch import Tensor\n+from torch.nn import Parameter\nfr... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | resolve import issue |
87,709 | 06.09.2021 11:37:04 | -3,600 | a259d91f6e3989f85d853895b69958f8ce74cab7 | relative path, not duplicate | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/recurrent/evolvegcnh.py",
"new_path": "torch_geometric_temporal/nn/recurrent/evolvegcnh.py",
"diff": "-import math\n-from typing import Optional, Tuple\n-\nimport torch\n-from torch import Tensor\n-from torch.nn import Parameter\nfr... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | relative path, not duplicate |
87,709 | 06.09.2021 13:29:28 | -3,600 | 7a0c892a42dcabb334f3d841e2fd4a2306feea96 | further improve coverage | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/recurrent/evolvegcno.py",
"new_path": "torch_geometric_temporal/nn/recurrent/evolvegcno.py",
"diff": "@@ -33,20 +33,6 @@ def gcn_norm(edge_index, edge_weight=None, num_nodes=None, improved=False,\nfill_value = 2. if improved else 1.... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | further improve coverage |
87,709 | 06.09.2021 15:29:19 | -3,600 | 38caa4e0c0c6be1491e72563618b8cb5fb2e7404 | importing instead of repeating from PyG functions | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/recurrent/evolvegcno.py",
"new_path": "torch_geometric_temporal/nn/recurrent/evolvegcno.py",
"diff": "-import math\nfrom typing import Optional, Tuple\nimport torch\nfrom torch import Tensor\n-from torch.nn import Parameter\nfrom to... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | importing instead of repeating from PyG functions |
87,723 | 12.10.2021 12:25:00 | -7,200 | e640ffb43f2b2ae7193ab4161adb9044f5dca466 | minor fixes in test suite: added missing assertions | [
{
"change_type": "MODIFY",
"old_path": "test/dataset_test.py",
"new_path": "test/dataset_test.py",
"diff": "@@ -139,10 +139,10 @@ def test_wiki():\ndataset = loader.get_dataset()\nfor epoch in range(1):\nfor snapshot in dataset:\n- snapshot.edge_index.shape == (2, 27079)\n- snapshot.edge_attr.shape ... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | minor fixes in test suite: added missing assertions |
87,727 | 17.12.2021 19:23:46 | -28,800 | a349541fa83d8d8ec5082c58caf0f66c00b5a0d9 | update the link of A3T-GCN | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/recurrent/attentiontemporalgcn.py",
"new_path": "torch_geometric_temporal/nn/recurrent/attentiontemporalgcn.py",
"diff": "@@ -6,7 +6,7 @@ from torch_geometric.nn import GCNConv\nclass A3TGCN(torch.nn.Module):\nr\"\"\"An implementati... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | update the link of A3T-GCN |
87,727 | 20.12.2021 17:54:06 | -28,800 | d8c35795273bd030df6e7b1515ed2e6784b366e0 | update a3tgcn_example. | [
{
"change_type": "MODIFY",
"old_path": "examples/recurrent/a3tgcn_example.py",
"new_path": "examples/recurrent/a3tgcn_example.py",
"diff": "@@ -14,18 +14,18 @@ dataset = loader.get_dataset()\ntrain_dataset, test_dataset = temporal_signal_split(dataset, train_ratio=0.2)\nclass RecurrentGCN(torch.nn.M... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | update a3tgcn_example. |
87,730 | 20.12.2021 14:26:14 | -3,600 | 032acfe18052e0171c1724c03f8b5c28c2704283 | Fixed EvolveGCN weight squeezing
Made squeeze dimension explicit, in order to avoid collapsing all dimensions for in_channels=1 | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/recurrent/evolvegcnh.py",
"new_path": "torch_geometric_temporal/nn/recurrent/evolvegcnh.py",
"diff": "@@ -95,5 +95,5 @@ class EvolveGCNH(torch.nn.Module):\nself.weight = self.initial_weight.data\nW = self.weight[None, :, :]\nX_tilde... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | Fixed EvolveGCN weight squeezing
Made squeeze dimension explicit, in order to avoid collapsing all dimensions for in_channels=1 |
87,724 | 26.12.2021 00:45:08 | -7,200 | ac3d2d6f869bfee6768fad408745e447ad99c271 | make attention weights trainable | [
{
"change_type": "MODIFY",
"old_path": "torch_geometric_temporal/nn/recurrent/attentiontemporalgcn.py",
"new_path": "torch_geometric_temporal/nn/recurrent/attentiontemporalgcn.py",
"diff": "@@ -24,7 +24,7 @@ class A3TGCN(torch.nn.Module):\nperiods: int,\nimproved: bool = False,\ncached: bool = False... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | make attention weights trainable |
87,724 | 26.12.2021 01:21:14 | -7,200 | 5fd5dbeea6dedb95e31ee40ae524824ad8c428c3 | add a new example for A3TGCN2 which supports batches | [
{
"change_type": "ADD",
"old_path": null,
"new_path": "examples/recurrent/a3tgcn2_example.py",
"diff": "+# I published a working notebook of this example at https://www.kaggle.com/elmahy/a3t-gcn-for-traffic-forecasting\n+\n+# The contribution makes training possible because it support batches of dat... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | add a new example for A3TGCN2 which supports batches |
87,724 | 26.12.2021 14:10:15 | -7,200 | 703a3541f46bb54ad2b893353644968b38a0bd9f | fix the version of torch_geometric at 1.7 to avoid the error of UnintializedParameter attribute not found | [
{
"change_type": "MODIFY",
"old_path": "setup.py",
"new_path": "setup.py",
"diff": "@@ -7,7 +7,7 @@ install_requires = [\n\"torch_scatter\",\n\"torch_cluster\",\n\"torch_spline_conv\",\n- \"torch_geometric\",\n+ \"torch_geometric==1.7.0\",\n\"numpy\",\n\"scipy\",\n\"tqdm\",\n"
},
{
"change_t... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | fix the version of torch_geometric at 1.7 to avoid the error of UnintializedParameter attribute not found |
87,724 | 26.12.2021 14:26:06 | -7,200 | e5dddd86e28ce35312de1cc82607a733db22fdbc | fix a typo as torch.nn instead of nn | [
{
"change_type": "MODIFY",
"old_path": "test/recurrent_test.py",
"new_path": "test/recurrent_test.py",
"diff": "@@ -23,6 +23,8 @@ def create_mock_data(number_of_nodes, edge_per_node, in_channels):\nreturn X, edge_index\n+\n+\ndef create_mock_attention_data(number_of_nodes, edge_per_node, in_channels... | Python | MIT License | benedekrozemberczki/pytorch_geometric_temporal | fix a typo as torch.nn instead of nn |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.