forex/Threshold.py at main · eachonly/forex · 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
import pandas as pd
import numpy as np
import math

# gets upper bound for the threshold


def get_upper_threshold(close):
    difference = close.diff()
    difference[0] = 0
    difference = difference.abs()

    bins = pd.cut(difference, bins=10)

    bins = bins.value_counts().to_frame().reset_index()
    bins["index"] = bins["index"].apply(lambda x: x.right)

    bins = bins.to_numpy()

    percentile_count = len(difference) * 0.85

    count = 0
    for i in range(10):
        count += bins[i, 1]
        if count > percentile_count:
            return bins[i, 0]

# calculate entropy


def get_entropy(labels, base=None):
    vc = pd.Series(labels).value_counts(normalize=True, sort=False)
    base = math.e if base is None else base
    return -(vc * np.log(vc)/np.log(base)).sum()

# get best threshold


def get_threshold(close):
    difference = close.diff()
    difference = difference.drop(0)
    difference = difference.tolist()

    threshold = 0
    thres_upper_bound = get_upper_threshold(close)
    temp_thres = 0
    best_entropy = -float('inf')

    while temp_thres < thres_upper_bound:
        labels = []
        for diff in difference:
            if diff > temp_thres:
                labels.append(2)
            elif -diff > temp_thres:
                labels.append(1)
            else:
                labels.append(0)
        entropy = get_entropy(labels)
        if entropy > best_entropy:
            best_entropy = entropy
            threshold = temp_thres
        temp_thres = temp_thres + 0.00001
    return threshold


if __name__ == "main":
    dataframe = pd.read_csv("data.csv")
    close = dataframe["Close"]