File size: 9,612 Bytes
6dd3ebe
 
 
 
 
 
4cc7625
 
 
6dd3ebe
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.fft
import math

# Modified from: https://github.com/thuml/Time-Series-Library
# Modified by Shourya Bose, shbose@ucsc.edu

class Inception_Block_V1(nn.Module):
    def __init__(self, in_channels, out_channels, num_kernels=6, init_weight=True):
        super(Inception_Block_V1, self).__init__()
        self.in_channels = in_channels
        self.out_channels = out_channels
        self.num_kernels = num_kernels
        kernels = []
        for i in range(self.num_kernels):
            kernels.append(nn.Conv2d(in_channels, out_channels, kernel_size=2 * i + 1, padding=i))
        self.kernels = nn.ModuleList(kernels)
        if init_weight:
            self._initialize_weights()

    def _initialize_weights(self):
        for m in self.modules():
            if isinstance(m, nn.Conv2d):
                nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu')
                if m.bias is not None:
                    nn.init.constant_(m.bias, 0)

    def forward(self, x):
        res_list = []
        for i in range(self.num_kernels):
            res_list.append(self.kernels[i](x))
        res = torch.stack(res_list, dim=-1).mean(-1)
        return res

class PositionalEmbedding(nn.Module):
    def __init__(self, d_model, max_len=5000):
        super(PositionalEmbedding, self).__init__()
        # Compute the positional encodings once in log space.
        pe = torch.zeros(max_len, d_model).float()
        pe.require_grad = False

        position = torch.arange(0, max_len).float().unsqueeze(1)
        div_term = (torch.arange(0, d_model, 2).float()
                    * -(math.log(10000.0) / d_model)).exp()

        pe[:, 0::2] = torch.sin(position * div_term)
        pe[:, 1::2] = torch.cos(position * div_term)

        pe = pe.unsqueeze(0)
        self.register_buffer('pe', pe)

    def forward(self, x):
        return self.pe[:, :x.size(1)]

class FixedEmbedding(nn.Module):
    def __init__(self, c_in, d_model):
        super(FixedEmbedding, self).__init__()

        w = torch.zeros(c_in, d_model).float()
        w.require_grad = False

        position = torch.arange(0, c_in).float().unsqueeze(1)
        div_term = (torch.arange(0, d_model, 2).float()
                    * -(math.log(10000.0) / d_model)).exp()

        w[:, 0::2] = torch.sin(position * div_term)
        w[:, 1::2] = torch.cos(position * div_term)

        self.emb = nn.Embedding(c_in, d_model)
        self.emb.weight = nn.Parameter(w, requires_grad=False)

    def forward(self, x):
        return self.emb(x).detach()

class TemporalEmbedding(nn.Module):
    def __init__(self, d_model, embed_type='fixed', freq='h'):
        super(TemporalEmbedding, self).__init__()

        hour_size = 96
        weekday_size = 7

        Embed = FixedEmbedding if embed_type == 'fixed' else nn.Embedding
        self.hour_embed = Embed(hour_size, d_model)
        self.weekday_embed = Embed(weekday_size, d_model)
        
    def forward(self, x):
        x = x.long()
        hour_x = self.hour_embed(x[:, :, 0])
        weekday_x = self.weekday_embed(x[:, :, 1])

        return hour_x + weekday_x

class TokenEmbedding(nn.Module):
    def __init__(self, c_in, d_model):
        super(TokenEmbedding, self).__init__()
        padding = 1 if torch.__version__ >= '1.5.0' else 2
        self.tokenConv = nn.Conv1d(in_channels=c_in, out_channels=d_model,
                                   kernel_size=3, padding=padding, padding_mode='circular', bias=False)
        for m in self.modules():
            if isinstance(m, nn.Conv1d):
                nn.init.kaiming_normal_(
                    m.weight, mode='fan_in', nonlinearity='leaky_relu')

    def forward(self, x):
        x = self.tokenConv(x.permute(0, 2, 1)).transpose(1, 2)
        return x

class DataEmbedding(nn.Module):
    def __init__(self, c_in, d_model, embed_type='fixed', freq='h', dropout=0.1):
        super(DataEmbedding, self).__init__()

        self.value_embedding = TokenEmbedding(c_in=c_in, d_model=d_model)
        self.position_embedding = PositionalEmbedding(d_model=d_model)
        self.temporal_embedding = TemporalEmbedding(d_model=d_model, embed_type=embed_type,
                                                    freq=freq)
        self.dropout = nn.Dropout(p=dropout)

    def forward(self, x, x_mark):
        if x_mark is None:
            x = self.value_embedding(x) + self.position_embedding(x)
        else:
            x = self.value_embedding(
                x) + self.temporal_embedding(x_mark) + self.position_embedding(x)
        return self.dropout(x)

def FFT_for_Period(x, k=2):
    # [B, T, C]
    xf = torch.fft.rfft(x, dim=1)
    # find period by amplitudes
    frequency_list = abs(xf).mean(0).mean(-1)
    frequency_list[0] = 0
    _, top_list = torch.topk(frequency_list, k)
    top_list = top_list.detach().cpu().numpy()
    period = x.shape[1] // top_list
    return period, abs(xf).mean(-1)[:, top_list]


class TimesBlock(nn.Module):
    def __init__(self, seq_len, pred_len, top_k, d_model, d_ff, num_kernels):
        super(TimesBlock, self).__init__()
        self.seq_len = seq_len
        self.pred_len = pred_len
        self.k = top_k
        # parameter-efficient design
        self.conv = nn.Sequential(
            Inception_Block_V1(d_model, d_ff,
                               num_kernels=num_kernels),
            nn.GELU(),
            Inception_Block_V1(d_ff, d_model,
                               num_kernels=num_kernels)
        )

    def forward(self, x):
        B, T, N = x.size()
        period_list, period_weight = FFT_for_Period(x, self.k)

        res = []
        for i in range(self.k):
            period = period_list[i]
            # padding
            if (self.seq_len + self.pred_len) % period != 0:
                length = (
                                 ((self.seq_len + self.pred_len) // period) + 1) * period
                padding = torch.zeros([x.shape[0], (length - (self.seq_len + self.pred_len)), x.shape[2]]).to(x.device)
                out = torch.cat([x, padding], dim=1)
            else:
                length = (self.seq_len + self.pred_len)
                out = x
            # reshape
            out = out.reshape(B, length // period, period,
                              N).permute(0, 3, 1, 2).contiguous()
            # 2D conv: from 1d Variation to 2d Variation
            out = self.conv(out)
            # reshape back
            out = out.permute(0, 2, 3, 1).reshape(B, -1, N)
            res.append(out[:, :(self.seq_len + self.pred_len), :])
        res = torch.stack(res, dim=-1)
        # adaptive aggregation
        period_weight = F.softmax(period_weight, dim=1)
        period_weight = period_weight.unsqueeze(
            1).unsqueeze(1).repeat(1, T, N, 1)
        res = torch.sum(res * period_weight, -1)
        # residual connection
        res = res + x
        return res


class TimesNet(nn.Module):
    """
    Paper link: https://openreview.net/pdf?id=ju_Uqw384Oq
    """

    def __init__(
        self,
        enc_in,
        dec_in,
        c_out,
        pred_len,
        seq_len,
        output_attention = False,
        data_idx = [0,3,4,5,6,7],
        time_idx = [1,2],
        d_model = 16,
        d_ff = 64,
        e_layers = 2,
        top_k = 5,
        num_kernels = 2,
        dropout = 0.1
    ):
        super(TimesNet, self).__init__()

        self.data_idx = data_idx
        self.time_idx = time_idx
        self.dec_in = dec_in

        self.seq_len = seq_len
        self.pred_len = pred_len
        self.model = nn.ModuleList([TimesBlock(seq_len, pred_len, top_k, d_model, d_ff, num_kernels)
                                    for _ in range(e_layers)])
        self.enc_embedding = DataEmbedding(enc_in, d_model, 'fixed', 'h',
                                           dropout)
        self.layer = e_layers
        self.layer_norm = nn.LayerNorm(d_model)
        self.predict_linear = nn.Linear(
            self.seq_len, self.pred_len + self.seq_len)
        self.projection = nn.Linear(
            d_model, c_out, bias=True)

    def forecast(self, x_enc, x_mark_enc, x_dec, x_mark_dec):
        # Normalization from Non-stationary Transformer
        means = x_enc.mean(1, keepdim=True).detach()
        x_enc = x_enc - means
        stdev = torch.sqrt(
            torch.var(x_enc, dim=1, keepdim=True, unbiased=False) + 1e-5)
        x_enc /= stdev

        # embedding
        enc_out = self.enc_embedding(x_enc, x_mark_enc)  # [B,T,C]
        enc_out = self.predict_linear(enc_out.permute(0, 2, 1)).permute(
            0, 2, 1)  # align temporal dimension
        # TimesNet
        for i in range(self.layer):
            enc_out = self.layer_norm(self.model[i](enc_out))
        # porject back
        dec_out = self.projection(enc_out)

        # De-Normalization from Non-stationary Transformer
        dec_out = dec_out * \
                  (stdev[:, 0, :].unsqueeze(1).repeat(
                      1, self.pred_len + self.seq_len, 1))
        dec_out = dec_out + \
                  (means[:, 0, :].unsqueeze(1).repeat(
                      1, self.pred_len + self.seq_len, 1))
        return dec_out

    def forward(self, x, fut_time):
        
        x_enc = x[:,:,self.data_idx]
        x_mark_enc = x[:,:,self.time_idx]
        x_dec = torch.zeros((fut_time.shape[0],fut_time.shape[1],self.dec_in),dtype=fut_time.dtype,device=fut_time.device)
        x_mark_dec = fut_time
        
        return self.forecast(x_enc,x_mark_enc,x_dec,x_mark_dec)[:,-1,[0]]