apt/analyze.py at main · Advancing-Machine-Human-Reasoning-Lab/apt · 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
from pprint import pprint


def get_checks(token):
    ret = []
    is_first = True
    start_time = 0
    with open("sentences/checks", "r") as f:
        for l in f.readlines():
            l = l.split("\t")
            if l[0] == token:
                if is_first:
                    start_time = float(l[1])
                    is_first = False
                ret.append(
                    {
                        "timestamp": float(l[1]) - start_time,
                        "time_taken": float(l[2]),
                        "dataset": l[3],
                        "index": int(l[4]),
                        "sentence": l[5],
                        "candidate": l[6],
                        "bleurt": float(l[7]),
                        "mi": int(l[8]),
                        "dollars": float(l[9]),
                    }
                )
    return ret


def get_submits(token):
    ret = []
    is_first = True
    start_time = 0
    with open("sentences/submits", "r") as f:
        for l in f.readlines():
            try:
                l = l.split("\t")
                if l[0] == token:
                    if is_first:
                        start_time = float(l[1])
                        is_first = False
                    ret.append(
                        {
                            "timestamp": float(l[1]) - start_time,
                            "time_taken": float(l[2]),
                            "dataset": l[3],
                            "index": int(l[4]),
                            "sentence": l[5],
                            "candidate": l[6],
                            "bleurt": float(l[7]),
                            "mi": int(l[8]),
                            "dollars": float(l[9]),
                        }
                    )
            except:
                print(l)
    return ret


def print_sent_cand(d):  # a list of dict of sentences and candidates
    for r in d:
        pprint(
            {
                "sentence": r["sentence"],
                "candidate": r["candidate"],
                "bleurt": r["bleurt"],
            }
        )


# person = get_submits('BVZ2UNUUSH')
# print(len(person), person[-1]['timestamp']/60, sum(_['dollars'] for _ in person))
# print_sent_cand(person)

F = open("sentences/final_submits", "w+")
F.write(
    "\t".join(
        [
            "token",
            "time",
            "duration",
            "dataset",
            "index",
            "sentence",
            "candidate",
            "BLEURT",
            "MI",
            "dollar",
        ]
    )
    + "\n"
)
with open("sentences/submits", "r") as f:
    for l in f.readlines():
        l = l.split("\t")
        try:
            q = float(l[1])
            if l[0] != "NFIPX9BDAS" and len(l[0]) == 10:
                for _ in range(len(l)):
                    if l[_] == "mrpc":
                        l[_] = "msrp"
                    if l[_] == "True":
                        l[_] = "1"
                    if l[_] == "False":
                        l[_] = "0"
                F.write("\t".join(l))
        except:
            continue