-
Notifications
You must be signed in to change notification settings - Fork 1
/
dataset.py
154 lines (128 loc) · 8.34 KB
/
dataset.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
import pandas as pd
from tqdm import tqdm
from torch.utils.data import Dataset, DataLoader
class TrajectoryTrainDataset(Dataset):
def __init__(self, data_df, map_set, n_time_slot):
user_id2idx_dict, POI_id2idx_dict, cat_id2idx_dict = map_set
self.trajectories = {}
self.labels = {}
data_df = data_df.groupby(['trajectory_id']).filter(lambda x: len(x) > 2)
data_df['local_time'] = pd.to_datetime(data_df['local_time']) # convert time column to datetime format
data_df = data_df.sort_values(['user_id', 'local_time']) # sort
for trajectory_id, trajectory in tqdm(data_df.groupby('trajectory_id'), desc=f"Prepare training dataset"):
user_id = trajectory_id.split('_')[0]
user_idx = user_id2idx_dict[user_id]
traj_idx = len(self.trajectories)
self.trajectories[traj_idx] = []
self.labels[traj_idx] = []
cur_day_of_year = trajectory.iloc[0]['local_time'].day_of_year
self.trajectories[traj_idx].append([[] for _ in range(n_time_slot)])
for index in range(len(trajectory) - 1):
_, pid, cid, _, _, _, _, _, _, tim, _, _, _, _, _, coo = trajectory.iloc[index]
_, next_pid, next_cid, _, _, _, _, _, _, next_tim, _, _, _, _, _, next_coo = trajectory.iloc[index + 1]
POI_idx, cat_idx = POI_id2idx_dict[pid], cat_id2idx_dict[cid]
next_POI_idx, next_cat_idx = POI_id2idx_dict[next_pid], cat_id2idx_dict[next_cid]
features = [user_idx, POI_idx, cat_idx, coo]
tim_info = tim.hour * 4 + int(tim.minute / 15) # Divide the time into time zones with 15-min intervals
labels = [next_POI_idx, next_cat_idx, next_coo]
checkin = {'features': features, 'time': tim_info, 'labels': labels}
if next_tim.day_of_year != tim.day_of_year or index == len(trajectory) - 2:
self.labels[traj_idx].append(labels)
if tim.day_of_year == cur_day_of_year:
self.trajectories[traj_idx][-1][int(tim.hour / (24 / n_time_slot))].append(checkin)
else:
cur_day_of_year = tim.day_of_year
self.trajectories[traj_idx].append([[] for _ in range(n_time_slot)])
self.trajectories[traj_idx][-1][int(tim.hour / (24 / n_time_slot))].append(checkin)
print(f"Train dataset length: ", len(self.trajectories))
def __len__(self):
return len(self.trajectories)
def __getitem__(self, item):
return self.trajectories[item], self.labels[item]
class TrajectoryValDataset(Dataset):
def __init__(self, data_df, map_set, n_time_slot):
user_id2idx_dict, POI_id2idx_dict, cat_id2idx_dict = map_set
self.trajectories = {}
self.labels = {}
data_df['user_id'] = data_df['user_id'].astype(str)
data_df = data_df[data_df['user_id'].isin(user_id2idx_dict.keys())]
data_df = data_df[data_df['POI_id'].isin(POI_id2idx_dict.keys())] # Do the same as GETNext
data_df = data_df[data_df['POI_catid'].isin(cat_id2idx_dict.keys())]
data_df = data_df.groupby(['trajectory_id']).filter(lambda x: len(x) > 2)
data_df['local_time'] = pd.to_datetime(data_df['local_time']) # convert time column to datetime format
data_df = data_df.sort_values(['user_id', 'local_time']) # sort
for trajectory_id, trajectory in tqdm(data_df.groupby('trajectory_id'), desc=f"Prepare validation dataset"):
user_id = trajectory_id.split('_')[0]
user_idx = user_id2idx_dict[user_id]
traj_idx = len(self.trajectories)
self.trajectories[traj_idx] = []
self.labels[traj_idx] = []
cur_day_of_year = trajectory.iloc[0]['local_time'].day_of_year
self.trajectories[traj_idx].append([[] for _ in range(n_time_slot)])
for index in range(len(trajectory) - 1):
_, pid, cid, _, _, _, _, _, _, tim, _, _, _, _, _, coo = trajectory.iloc[index]
_, next_pid, next_cid, _, _, _, _, _, _, next_tim, _, _, _, _, _, next_coo = trajectory.iloc[index + 1]
POI_idx, cat_idx = POI_id2idx_dict[pid], cat_id2idx_dict[cid]
next_POI_idx, next_cat_idx = POI_id2idx_dict[next_pid], cat_id2idx_dict[next_cid]
features = [user_idx, POI_idx, cat_idx, coo]
tim_info = tim.hour * 4 + int(tim.minute / 15) # Divide the time into time zones with 15-min intervals
labels = [next_POI_idx, next_cat_idx, next_coo]
checkin = {'features': features, 'time': tim_info, 'labels': labels}
if next_tim.day_of_year != tim.day_of_year or index == len(trajectory) - 2:
self.labels[traj_idx].append(labels)
if tim.day_of_year == cur_day_of_year:
self.trajectories[traj_idx][-1][int(tim.hour / (24 / n_time_slot))].append(checkin)
else:
cur_day_of_year = tim.day_of_year
self.trajectories[traj_idx].append([[] for _ in range(n_time_slot)])
self.trajectories[traj_idx][-1][int(tim.hour / (24 / n_time_slot))].append(checkin)
print(f"Validation dataset length: ", len(self.trajectories))
def __len__(self):
return len(self.trajectories)
def __getitem__(self, item):
return self.trajectories[item], self.labels[item]
class TrajectoryTestDataset(Dataset):
def __init__(self, data_df, map_set, n_time_slot):
user_id2idx_dict, POI_id2idx_dict, cat_id2idx_dict = map_set
self.trajectories = {}
self.labels = {}
data_df['user_id'] = data_df['user_id'].astype(str)
data_df = data_df[data_df['user_id'].isin(user_id2idx_dict.keys())]
data_df = data_df[data_df['POI_id'].isin(POI_id2idx_dict.keys())] # Do the same as GETNext
data_df = data_df[data_df['POI_catid'].isin(cat_id2idx_dict.keys())]
data_df = data_df.groupby(['trajectory_id']).filter(lambda x: len(x) > 2)
data_df['local_time'] = pd.to_datetime(data_df['local_time']) # convert time column to datetime format
data_df = data_df.sort_values(['user_id', 'local_time']) # sort
for trajectory_id, trajectory in tqdm(data_df.groupby('trajectory_id'), desc=f"Prepare testing dataset"):
user_id = trajectory_id.split('_')[0]
user_idx = user_id2idx_dict[user_id]
traj_idx = len(self.trajectories)
self.trajectories[traj_idx] = []
self.labels[traj_idx] = []
cur_day_of_year = trajectory.iloc[0]['local_time'].day_of_year
self.trajectories[traj_idx].append([[] for _ in range(n_time_slot)])
for index in range(len(trajectory) - 1):
_, pid, cid, _, _, _, _, _, _, tim, _, _, _, _, _, coo = trajectory.iloc[index]
_, next_pid, next_cid, _, _, _, _, _, _, next_tim, _, _, _, _, _, next_coo = trajectory.iloc[index + 1]
POI_idx, cat_idx = POI_id2idx_dict[pid], cat_id2idx_dict[cid]
next_POI_idx, next_cat_idx = POI_id2idx_dict[next_pid], cat_id2idx_dict[next_cid]
features = [user_idx, POI_idx, cat_idx, coo]
tim_info = tim.hour * 4 + int(tim.minute / 15) # Divide the time into time zones with 15-min intervals
if index == len(trajectory) - 2: # test node
labels = [next_POI_idx, next_cat_idx, next_coo]
else:
labels = [-1, -1, -1]
checkin = {'features': features, 'time': tim_info, 'labels': labels}
if next_tim.day_of_year != tim.day_of_year or index == len(trajectory) - 2:
self.labels[traj_idx].append(labels)
if tim.day_of_year == cur_day_of_year:
self.trajectories[traj_idx][-1][int(tim.hour / (24 / n_time_slot))].append(checkin)
else:
cur_day_of_year = tim.day_of_year
self.trajectories[traj_idx].append([[] for _ in range(n_time_slot)])
self.trajectories[traj_idx][-1][int(tim.hour / (24 / n_time_slot))].append(checkin)
print(f"Test dataset length: ", len(self.trajectories))
def __len__(self):
return len(self.trajectories)
def __getitem__(self, item):
return self.trajectories[item], self.labels[item]