This repository has been archived by the owner on Jun 2, 2023. It is now read-only.
-
Notifications
You must be signed in to change notification settings - Fork 14
/
Copy pathSnakefile_rgcn_hypertune.smk
210 lines (189 loc) · 7.15 KB
/
Snakefile_rgcn_hypertune.smk
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
import os
from river_dl.preproc_utils import asRunConfig
from river_dl.preproc_utils import prep_all_data
from river_dl.torch_utils import train_torch
from river_dl.torch_utils import rmse_masked
from river_dl.evaluate import combined_metrics
import numpy as np
import torch
import torch.optim as optim
from river_dl.torch_models import RGCN_v1
from river_dl.predict import predict_from_io_data
out_dir = config['out_dir']
code_dir = ''
### Set up your hyperparameters as wild cards
offsets = [15,.25, .5, 1]
sequence_length = [60,180,365]
run_id = ['RGCN_Hypertune']
rule all:
input:
expand("{outdir}/{offset}_{seq_length}/{run_id}/{metric_type}_metrics.csv",
outdir=out_dir,
offset=offsets,
seq_length=sequence_length,
metric_type=['overall', 'month', 'reach'],
run_id=run_id
),
expand("{outdir}/asRunConfig_{run_id}.yml", outdir=out_dir, run_id = run_id),
expand("{outdir}/Snakefile_{run_id}", outdir=out_dir, run_id = run_id),
rule as_run_config:
output:
"{outdir}/asRunConfig_{run_id}.yml"
group: "prep"
run:
asRunConfig(config,code_dir,output[0])
rule copy_snakefile:
output:
"{outdir}/Snakefile_{run_id}"
group: "prep"
shell:
"""
scp Snakefile {output[0]}
"""
rule prep_io_data:
input:
config['sntemp_file'],
config['obs_file'],
config['dist_matrix_file'],
output:
"{outdir}/{offset}_{seq_length}/prepped.npz"
threads: 2
group: "prep"
run:
prep_all_data(
x_data_file=input[0],
pretrain_file=input[0],
y_data_file=input[1],
distfile=input[2],
x_vars=config['x_vars'],
y_vars_pretrain=config['y_vars_pretrain'],
y_vars_finetune=config['y_vars_finetune'],
catch_prop_file=None,
train_start_date=config['train_start_date'],
train_end_date=config['train_end_date'],
val_start_date=config['val_start_date'],
val_end_date=config['val_end_date'],
test_start_date=config['test_start_date'],
test_end_date=config['test_end_date'],
segs=None,
out_file=output[0],
trn_offset= float(wildcards.offset),
tst_val_offset= float(wildcards.offset),
seq_len=int(wildcards.seq_length),
)
# Pretrain the model on process based model
rule pre_train:
input:
"{outdir}/{offset}_{seq_length}/prepped.npz"
output:
"{outdir}/{offset}_{seq_length}/{run_id}/pretrained_weights.pth",
"{outdir}/{offset}_{seq_length}/{run_id}/pretrain_log.csv",
threads: 4
group: 'train'
run:
os.system("module load analytics cuda11.3/toolkit/11.3.0")
os.system("export LD_LIBRARY_PATH=/cm/shared/apps/nvidia/TensorRT-6.0.1.5/lib:/cm/shared/apps/nvidia/cudnn_8.0.5/lib64:$LD_LIBRARY_PATH")
data = np.load(input[0])
num_segs = len(np.unique(data['ids_trn']))
adj_mx = data['dist_matrix']
in_dim = len(data['x_vars'])
device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')
model = RGCN_v1(in_dim,config['hidden_size'],adj_mx,device=device)
opt = optim.Adam(model.parameters(),lr=config['pretrain_learning_rate'])
train_torch(model,
loss_function=rmse_masked,
optimizer=opt,
x_train=data['x_pre_full'],
y_train=data['y_pre_full'],
max_epochs=config['pt_epochs'],
batch_size=num_segs,
weights_file=output[0],
log_file=output[1],
device=device,
keep_portion=float(wildcards.offset))
# Finetune/train the model on observations
rule finetune_train:
input:
"{outdir}/{offset}_{seq_length}/prepped.npz",
"{outdir}/{offset}_{seq_length}/{run_id}/pretrained_weights.pth",
"{outdir}/{offset}_{seq_length}/{run_id}/pretrain_log.csv",
output:
"{outdir}/{offset}_{seq_length}/{run_id}/finetuned_weights.pth",
"{outdir}/{offset}_{seq_length}/{run_id}/finetune_log.csv",
threads: 4
group: 'train'
run:
os.system("module load analytics cuda11.3/toolkit/11.3.0")
os.system("export LD_LIBRARY_PATH=/cm/shared/apps/nvidia/TensorRT-6.0.1.5/lib:/cm/shared/apps/nvidia/cudnn_8.0.5/lib64:$LD_LIBRARY_PATH")
data = np.load(input[0])
num_segs = len(np.unique(data['ids_trn']))
adj_mx = data['dist_matrix']
in_dim = len(data['x_vars'])
device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')
model = RGCN_v1(in_dim,config['hidden_size'],adj_mx,device=device)
opt = optim.Adam(model.parameters(),lr=config['finetune_learning_rate'])
scheduler = optim.lr_scheduler.LambdaLR(opt,lr_lambda=lambda epoch: 0.97 ** epoch)
model.load_state_dict(torch.load(input[1]))
train_torch(model,
loss_function=rmse_masked,
optimizer=opt,
x_train=data['x_trn'],
y_train=data['y_obs_trn'],
x_val=data['x_val'],
y_val=data['y_obs_val'],
max_epochs=config['ft_epochs'],
early_stopping_patience=config['early_stopping'],
batch_size=num_segs,
weights_file=output[0],
log_file=output[1],
device=device,
keep_portion=float(wildcards.offset))
rule make_predictions:
input:
"{outdir}/{offset}_{seq_length}/{run_id}/finetuned_weights.pth",
"{outdir}/{offset}_{seq_length}/prepped.npz"
output:
"{outdir}/{offset}_{seq_length}/{run_id}/{partition}_preds.feather",
group: 'train'
threads: 3
run:
data = np.load(input[1])
adj_mx = data['dist_matrix']
in_dim = len(data['x_vars'])
model = RGCN_v1(in_dim,config['hidden_size'],adj_mx)
opt = optim.Adam(model.parameters(),lr=config['finetune_learning_rate'])
model.load_state_dict(torch.load(input[0]))
predict_from_io_data(model=model,
io_data=input[1],
partition=wildcards.partition,
outfile=output[0],
trn_offset=float(wildcards.offset),
tst_val_offset=float(wildcards.offset))
def get_grp_arg(wildcards):
if wildcards.metric_type == 'overall':
return None
elif wildcards.metric_type == 'month':
return 'month'
elif wildcards.metric_type == 'reach':
return 'seg_id_nat'
elif wildcards.metric_type == 'month_reach':
return ['seg_id_nat', 'month']
rule combine_metrics:
input:
config['obs_file'],
"{outdir}/{offset}_{seq_length}/{run_id}/trn_preds.feather",
"{outdir}/{offset}_{seq_length}/{run_id}/val_preds.feather",
"{outdir}/{offset}_{seq_length}/{run_id}/tst_preds.feather"
output:
"{outdir}/{offset}_{seq_length}/{run_id}/{metric_type}_metrics.csv"
group: 'predict_eval'
threads: 3
params:
grp_arg=get_grp_arg
run:
combined_metrics(obs_file=input[0],
pred_trn=input[1],
pred_val=input[2],
pred_tst=input[3],
group=params.grp_arg,
outfile=output[0])