GnnForPrivacyScan/train/model.py

import torch
import torch.nn as nn
import torch.nn.init as init


class AttrProxy(object):
    """
    Translates index lookups into attribute lookups.
    To implement some trick which able to use list of nn.Module in a nn.Module
    see https://discuss.pytorch.org/t/list-of-nn-module-in-a-nn-module/219/2
    """
    def __init__(self, module, prefix):
        self.module = module
        self.prefix = prefix

    def __getitem__(self, i):
        return getattr(self.module, self.prefix + str(i))


class Propogator(nn.Module):
    """
    Gated Propogator for GGNN
    Using LSTM gating mechanism
    """

    def __init__(self, state_dim, n_node, n_edge_types):
        super(Propogator, self).__init__()

        self.n_node = n_node
        self.n_edge_types = n_edge_types

        self.reset_gate = nn.Sequential(
            nn.Linear(state_dim * 3, state_dim),
            nn.Sigmoid()
        )
        self.update_gate = nn.Sequential(
            nn.Linear(state_dim * 3, state_dim),
            nn.Sigmoid()
        )
        self.tansform = nn.Sequential(
            nn.Linear(state_dim * 3, state_dim),
            nn.Tanh()
        )

    def forward(self, state_in, state_out, state_cur, A):
        A_in = A[:, :, :self.n_node * self.n_edge_types]
        A_out = A[:, :, self.n_node * self.n_edge_types:]

        a_in = torch.bmm(A_in, state_in)
        a_out = torch.bmm(A_out, state_out)
        a = torch.cat((a_in, a_out, state_cur), 2)

        r = self.reset_gate(a)
        z = self.update_gate(a)

        joined_input = torch.cat((a_in, a_out, r * state_cur), 2)
        h_hat = self.tansform(joined_input)

        output = (1 - z) * state_cur + z * h_hat

        return output


class GGNN(nn.Module):
    """
    Gated Graph Sequence Neural Networks (GGNN)
    Mode: SelectNode
    Implementation based on https://arxiv.org/abs/1511.05493
    """

    def __init__(self, opt):
        super(GGNN, self).__init__()

        # assert (opt.state_dim >= opt.annotation_dim, 'state_dim must be no less than annotation_dim')
        self.is_training_ggnn = opt.is_training_ggnn
        self.state_dim = opt.state_dim
        self.n_edge_types = opt.n_edge_types
        self.n_node = opt.n_node
        self.n_steps = opt.n_steps
        self.n_classes = opt.n_classes

        for i in range(self.n_edge_types):
            # incoming and outgoing edge embedding
            in_fc = nn.Linear(self.state_dim, self.state_dim)
            out_fc = nn.Linear(self.state_dim, self.state_dim)
            self.add_module("in_{}".format(i), in_fc)
            self.add_module("out_{}".format(i), out_fc)

        self.in_fcs = AttrProxy(self, "in_")
        self.out_fcs = AttrProxy(self, "out_")

        # Propogation Model
        self.propogator = Propogator(self.state_dim, self.n_node, self.n_edge_types)

        # Output Model
        self.out = nn.Sequential(
            nn.Linear(self.state_dim, self.state_dim),
            nn.LeakyReLU(),
            nn.Linear(self.state_dim, 1),
            nn.Tanh(),
        )

        self.soft_attention = nn.Sequential(
            nn.Linear(self.state_dim, self.state_dim),
            nn.LeakyReLU(),
            nn.Linear(self.state_dim, 1),
            nn.Sigmoid(),

        )

        self.class_prediction = nn.Sequential(
            nn.Linear(opt.state_dim, opt.n_hidden),
            nn.LeakyReLU(),
            nn.Linear(opt.n_hidden, opt.n_classes),
            # nn.Softmax(dim=1)
        )

        # self.class_prediction = nn.Sequential(
        #     nn.Linear(opt.n_node, opt.n_classes),
        #     nn.Softmax(dim=1)
        # )

        self._initialization()

    def _initialization(self):
        for m in self.modules():
            if isinstance(m, nn.Linear):
                init.xavier_normal_(m.weight.data)
                if m.bias is not None:
                    init.normal_(m.bias.data)

    def forward(self, prop_state, A):
        # print(prop_state.shape)
        for i_step in range(self.n_steps):
            in_states = []
            out_states = []
            for i in range(self.n_edge_types):
                in_states.append(self.in_fcs[i](prop_state))
                out_states.append(self.out_fcs[i](prop_state))

            in_states = torch.stack(in_states).transpose(0, 1).contiguous()
            in_states = in_states.view(-1, self.n_node * self.n_edge_types, self.state_dim)

            out_states = torch.stack(out_states).transpose(0, 1).contiguous()
            out_states = out_states.view(-1, self.n_node * self.n_edge_types, self.state_dim)

            prop_state = self.propogator(in_states, out_states, prop_state, A)

        # print("Prop state : " + str(prop_state.shape))
        # output = self.out(prop_state)
        # print("Out : " + str(output.shape))

        soft_attention_ouput = self.soft_attention(prop_state)
        # print("Soft : " + str(soft_attention_ouput.shape))
        # Element wise hadamard product to get the graph representation, check Equation 7 in GGNN paper for more details
        output = torch.mul(prop_state, soft_attention_ouput)
        # print("Out : " + str(output.shape))
        output = output.sum(1)
        # print(output.shape)
        if self.is_training_ggnn == True:
            output = self.class_prediction(output)

        return output