SFL/eval.py at main · ispc-lab/SFL · GitHub

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
import numpy as np
import torch
from sklearn import metrics
import os
import argparse
from utils.load_utils import load_data, load_model
from utils.dataloader import get_data_loader
from utils.train import set_seed
import pandas as pd

def evaluate(args, dataloader, dir):
    with torch.no_grad():
        args.model.eval()
        y_pred = np.array([])
        y_true = np.array([])

        for batch in dataloader:
            batch = {k: v.to(args.device) for k, v in batch.items()}
            compound_class = batch['label']

            output, _ = args.model(batch)

            output_org = output * args.std + args.mean
            compound_class = compound_class * args.std + args.mean

            y_pred = np.concatenate((y_pred, output_org[:, 0].detach().cpu().numpy()))
            y_true = np.concatenate((y_true, compound_class.detach().cpu().numpy()))

        r2 = metrics.r2_score(y_true, y_pred)
        mae = metrics.mean_absolute_error(y_true, y_pred)
        rmse = np.sqrt(metrics.mean_squared_error(y_true, y_pred))

    test_df = pd.DataFrame({})
    test_df['label'] = y_true
    test_df['pred'] = y_pred

    save_pred_path = os.path.join(args.load_model_dir, os.path.join(dir, 'pred.csv'))

    test_df.to_csv(save_pred_path, index=False)
    return r2, mae, rmse


def build_argparse():
    parser = argparse.ArgumentParser()
    parser.add_argument("--use_SFL", default=1, type=int)
    parser.add_argument("--note", default=None, type=str)
    parser.add_argument('--arch_type', default='unimol', type=str,
                        help="unimol, unimol2_84M, schnet, egnn, dimenet++, visnet, attentive_fp")
    parser.add_argument('--dataset', default='qm7', type=str,
                        help="esol, freesolv, lipo, qm7")
    parser.add_argument('--save_dir', default='./ckpt/SFL', type=str)
    parser.add_argument('--device', default='cuda:0', type=str)
    parser.add_argument('--batch_size', default=256, type=int)
    parser.add_argument('--num_workers', default=7, type=int)

    args = parser.parse_args()
    args = load_data(args)
    args = load_model(args)

    if args.note:
        args.load_model_dir = os.path.join(args.save_dir, args.arch_type + '_' + args.note + '_' + args.dataset)
    else:
        args.load_model_dir = os.path.join(args.save_dir, args.arch_type + '_' + args.dataset)
    return args


if __name__ == '__main__':
    args = build_argparse()

    df_list = []
    for dir in os.listdir(args.load_model_dir):
        if not dir.endswith('.txt'):
            set_seed(int(dir))
            load_model_path = os.path.join(args.load_model_dir, os.path.join(dir, 'best_model.pt'))
            save_result_path = os.path.join(args.load_model_dir, os.path.join(dir, 'results.csv'))

            state_dict = torch.load(load_model_path, map_location=args.device)

            args.model.load_state_dict(state_dict)

            modes = ['test_xood', 'test_yood', 'test_xyood']

            results = {}
            for mode in modes:
                dataloader = get_data_loader(args, mode)
                results[mode] = evaluate(args, dataloader, dir)

            r2_xood, mae_xood, rmse_xood = results['test_xood']
            r2_yood, mae_yood, rmse_yood = results['test_yood']
            r2_xyood, mae_xyood, rmse_xyood = results['test_xyood']

            rmse_score = (rmse_xood + rmse_yood + rmse_xyood) / 3
            r2_score = (r2_xood + r2_yood + r2_xyood) / 3
            mae_score = (mae_xood + mae_yood + mae_xyood) / 3

            data = {
                'Metric': ['R2', 'MAE', 'RMSE'],
                'Score': [r2_score, mae_score, rmse_score],
                'XOOD': [r2_xood, mae_xood, rmse_xood],
                'YOOD': [r2_yood, mae_yood, rmse_yood],
                'XYOOD': [r2_xyood, mae_xyood, rmse_xyood],
            }

            df = pd.DataFrame(data)
            df.to_csv(save_result_path, index=False)

            df_list.append(df)

    df_combined = pd.concat(df_list)

    average_scores = df_combined.groupby('Metric').mean()
    std_scores = df_combined.groupby('Metric').std()

    formatted_scores = average_scores.applymap(lambda x: f"& {x:.6f}") + std_scores.applymap(lambda x: f"±{x:.6f}")

    formatted_string = formatted_scores.to_string(index=True, header=True)

    with open(os.path.join(args.load_model_dir, args.arch_type + '_' + args.dataset + '_results.txt'), 'w', encoding='utf-8') as file:
        file.write(formatted_string)

    print(formatted_string)