-
Notifications
You must be signed in to change notification settings - Fork 2
/
partner_selection.py
45 lines (33 loc) · 1.46 KB
/
partner_selection.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
'''
Module for the implementation of the Partner Selection algorithms.
'''
#import needed packages
import pandas as pd
import numpy as np
import spicy as sp
class PartnersSelection:
'''Class that contains algorithms that are used in partners selection.
'''
def __init__(self):
pass
def get_corr_matrix(self, returns_df ):
'''Get the Spearman's correlation matrix from the returns matrix.
parameters:(pandas df) returns_df
returns: (pandas df) df_corr
'''
df_corr = returns_df.corr("spearman", min_periods=1)
return df_corr
def employ_trad_approach(self,df_corr):
df_corr = df_corr.fillna(0)
df_corr = df_corr.set_index(keys=df_corr.columns)
df_corr_max = pd.DataFrame(df_corr.columns.values[np.argsort(-df_corr.values, axis=1)[:, :4]],
index=df_corr.index,
columns=['Self_Ticker', '1st Max', '2nd Max', '3rd Max'])
df_corr_max.drop('Self_Ticker', axis=1, inplace=True)
df_corr_sort = df_corr.apply(lambda x: np.sort(x), axis=1, raw=True)
df_corr_sort = df_corr_sort.iloc[:, -4:]
df_corr_sort = df_corr_sort.iloc[:, :-1]
df_corr_sort.loc[:, "Sum"] = df_corr_sort.sum(axis=1)
target_stock = df_corr_sort["Sum"].idxmax(axis=0)
partners = df_corr_max.loc[target_stock].tolist()
return (target_stock, partners)