-
Notifications
You must be signed in to change notification settings - Fork 0
/
feature_engineering.py
71 lines (58 loc) · 2.83 KB
/
feature_engineering.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
import pandas as pd
from sklearn.preprocessing import PolynomialFeatures
class FeatureEngineering():
def __init__(self, df: pd.DataFrame):
self.data = df.copy()
self.df_fe = pd.DataFrame()
self.target = 'price.close'
def derivada(self):
for column in self.data.columns:
if column != self.target:
self.df_fe[f'{column}_derivative'] = self.data[column].diff()
def integral(self):
for column in self.data.columns:
if column != self.target:
#Integral numa janela de 3 dias.
self.df_fe[f'{column}_integral'] = self.data[column].rolling(3).sum()
def momentos_estatisticos(self):
for column in self.data.columns:
if column != self.target:
#Média móvel e desvio padrão de 3 dias.
self.df_fe[f'{column}_moving_average'] = self.data[column].rolling(3).mean()
self.df_fe[f'{column}_std'] = self.data[column].rolling(3).std()
def combinacoes_polinomiais(self):
df_poly = PolynomialFeatures(2)
cols = self.data.columns
df_poly = pd.DataFrame(df_poly.fit_transform(self.data[cols]))
qtde_colunas = len(df_poly.columns)
df_poly = df_poly.drop(columns=[x for x in range (len(cols)+1)])
nome_novas_colunas = []
nao_vistadas = list(cols.copy())
for coluna in cols:
atual = coluna
nome_novas_colunas.append(f'{coluna}^2')
for _ in nao_vistadas:
if (_ != atual):
nome_novas_colunas.append(f'{coluna}*{_}')
nao_vistadas.remove(coluna)
nome_velhas_colunas = [x for x in range(len(cols)+1,qtde_colunas)]
for i in range(nome_velhas_colunas[0],nome_velhas_colunas[-1]+1):
df_poly = df_poly.rename(columns={i:nome_novas_colunas[i-nome_velhas_colunas[0]]})
for col in df_poly.columns:
self.df_fe[col] = df_poly[col].values
def difference(self):
df_final = pd.DataFrame()
df_final['high-low'] = self.data['price.high'] - self.data['price.low']
df_final['high-close'] = self.data['price.high'] - self.data['price.close']
df_final['low-close'] = self.data['price.low'] - self.data['price.close']
df_final['close-open'] = self.data['price.close'] - self.data['price.open']
df_final['high-open'] = self.data['price.high'] - self.data['price.open']
df_final['low-open'] = self.data['price.low'] - self.data['price.open']
self.df_fe = self.df_fe.merge(df_final, left_index=True, right_index=True)
def pipeline_feat_eng(self):
self.derivada()
self.integral()
self.momentos_estatisticos()
self.combinacoes_polinomiais()
self.difference()
return self.df_fe.copy()