def evaluation_method_3(self):
self.filedf = pd.read_csv('factsheetdf.csv',index_col='Unnamed: 0')
self.filedf.sort_values('r6posrate',inplace = True,ascending=False)
#take by posrate and negrate
upper = self.filedf[self.filedf['r6posrate'] >= 0.8]
upper = upper[upper['total_hits'] >= 10]
lower = self.filedf[self.filedf['r6negrate'] >= 0.8]
lower = lower[lower['total_hits'] >= 10]
temp = pd.concat([upper,lower])
#70% by viability
temp['r6difference'] = temp['r6posrate'] - temp['r6negrate']
temp['r6viable'] = temp['r6difference'] * temp['r6mean']
temp = temp[temp['r6viable'] > 0]
self.usabledf = temp
genes = temp.index.tolist()
pool_genes = []
for x in genes:
pool_genes.append(ast.literal_eval(x))
self.genes = pool_genes
df = pd.read_csv('df.csv')
_, self.binDict = qe.makeDf(df)
self.inverseGenes()
def evaluation_method_1(self):
self.filedf = pd.read_csv('factsheetdf.csv',index_col='Unnamed: 0')
self.filedf.sort_values('r5mean',inplace = True,ascending=False)
#take highest and lowest 30% return
how_many_to_take = int(len(self.filedf)*0.3)
temp = pd.concat([self.filedf.head(how_many_to_take),self.filedf.tail(how_many_to_take)])
temp['r5difference'] = temp['r5posrate'] - temp['r5negrate']
temp['r5viable'] = temp['r5difference'] * temp['r5mean']
temp = temp[temp['r5viable'] > 0]
temp.sort_values('total_hits',inplace=True,ascending=False)
temp = temp.tail(int(len(temp)*0.8))
temp.sort_values('r5mean',inplace=True,ascending=False)
temp['r5meanabs'] = abs(temp['r5mean'])
temp.sort_values('r5meanabs',inplace=True,ascending=False)
temp = temp[temp['total_hits'] > 10]
self.usabledf = temp
genes = temp.index.tolist()
pool_genes = []
for x in genes:
pool_genes.append(ast.literal_eval(x))
self.genes = pool_genes
df = pd.read_csv('df.csv')
_, self.binDict = qe.makeDf(df)
self.inverseGenes()
def evaluation_method_2(self):
self.filedf = pd.read_csv('factsheetdf.csv',index_col='Unnamed: 0')
self.filedf.sort_values('total_hits',inplace = True,ascending=False)
#take the upper 50%
temp = self.filedf.head(int(len(self.filedf)*0.5))
#take 50% by mean
temp.sort_values('r5mean',inplace = True,ascending=False)
how_many_to_take = int(len(self.filedf) * 0.25)
temp = pd.concat([temp.head(how_many_to_take),temp.tail(how_many_to_take)])
#70% by viability
temp['r5difference'] = temp['r5posrate'] - temp['r5negrate']
temp['r5viable'] = temp['r5difference'] * temp['r5mean']
temp.sort_values('r5viable',inplace = True,ascending=False)
temp = temp.head(int(len(temp) * 0.3))
self.usabledf = temp
genes = temp.index.tolist()
pool_genes = []
for x in genes:
pool_genes.append(ast.literal_eval(x))
self.genes = pool_genes
df = pd.read_csv('df.csv')
_, self.binDict = qe.makeDf(df)
self.inverseGenes()
def evaluation_method_5(self):
self.filedf = pd.read_csv('factsheetdf.csv', index_col='Unnamed: 0')
self.filedf.dropna(inplace=True)
self.filedf = self.filedf[self.filedf['total_hits'] >= 20]
temp = self.filedf
self.usabledf = temp
genes = temp.index.tolist()
pool_genes = []
for x in genes:
pool_genes.append(ast.literal_eval(x))
self.genes = pool_genes
df = pd.read_csv('df.csv')
_, self.binDict = qe.makeDf(df)
self.inverseGenes()
def evaluation_method_6(self):
self.filedf = pd.read_csv('factsheetdf.csv', index_col='Unnamed: 0')
self.filedf['r7-r2'] = self.filedf['r7mean'] - self.filedf['r2mean']
self.filedf[
'r7-r2samesign'] = self.filedf['r2mean'] * self.filedf['r7-r2'] > 0
#for those that are of the same sign, take the highest
self.filedf.sort_values('r7-r2', inplace=True, ascending=False)
upper = self.filedf.head(int(len(self.filedf) * 0.4))
upperdiffsign = upper[upper['r7-r2samesign'] == False]
uppersamesign = upper[upper['r7-r2samesign'] == True]
lower = self.filedf.tail(int(len(self.filedf) * 0.4))
lowersamesign = lower[lower['r7-r2samesign'] == True]
lowerdiffsign = lower[lower['r7-r2samesign'] == False]
diffsigntemp = pd.concat([upperdiffsign, lowerdiffsign])
samesigntemp = pd.concat([uppersamesign, lowersamesign])
diffsigntemp['r7divr2'] = (diffsigntemp['r7-r2']) / abs(
diffsigntemp['r2mean'])
samesigntemp['r7divr2'] = (
samesigntemp['r7-r2']) / samesigntemp['r2mean']
samesigntemp.sort_values('r7divr2', inplace=True, ascending=False)
diffsigntemp.sort_values('r7divr2', inplace=True, ascending=False)
#take 20% from them
samesigntemp = samesigntemp.head(int(len(samesigntemp) * 0.2))
diffsigntemp = diffsigntemp.head(int(len(diffsigntemp) * 0.2))
temp = pd.concat([samesigntemp, diffsigntemp])
temp['r7-r2viable'] = temp['r7-r2'] * (temp['r7posrate'] -
temp['r7negrate'])
temp.sort_values('r7-r2viable', inplace=True, ascending=False)
temp = temp.head(int(len(temp) * 0.8))
self.usabledf = temp
genes = temp.index.tolist()
pool_genes = []
for x in genes:
pool_genes.append(ast.literal_eval(x))
self.genes = pool_genes
df = pd.read_csv('df.csv')
_, self.binDict = qe.makeDf(df)
self.inverseGenes()
# -*- coding: utf-8 -*-
import pandas as pd
import quickEngine
import numpy as np
import geneGenerator
from multiprocessing import Process, Manager, Pool
from concurrent.futures import ProcessPoolExecutor
import numpy as np
import time
import multiprocessing
import multiprocessing.pool
df = pd.read_csv('df.csv')
column_index_value_dict, binDict = quickEngine.makeDf(df)
generator = geneGenerator.geneGenerator(5,10,20)
'''
genes = generator.generateMany(100)
factsheetdict = evaluategenes(genes,column_index_value_dict)
factsheetdf = pd.DataFrame.from_dict(factsheetdict,orient='index',columns=['hits_per_day','total_hits','r2mean','r2posrate','r2negrate','r3mean','r3posrate','r3negrate','r4mean','r4posrate','r4negrate','r5mean','r5posrate','r5negrate','r6mean','r6posrate','r6negrate','r7mean','r7posrate','r7negrate','r8mean','r8posrate','r8negrate','r9mean','r9posrate','r9negrate','r10mean','r10posrate','r10negrate'])
genes = generator.generateMany(1000)
factsheetdict = evaluategenes(genes,column_index_value_dict)
factsheetdf = pd.DataFrame.from_dict(factsheetdict,orient='index',columns=['hits_per_day','total_hits','r2mean','r2posrate','r2negrate','r3mean','r3posrate','r3negrate','r4mean','r4posrate','r4negrate','r5mean','r5posrate','r5negrate','r6mean','r6posrate','r6negrate','r7mean','r7posrate','r7negrate','r8mean','r8posrate','r8negrate','r9mean','r9posrate','r9negrate','r10mean','r10posrate','r10negrate'])
def evaluation_method_4(self):
#First step is to get lowest 30% for all r1
self.filedf = pd.read_csv('factsheetdf.csv', index_col='Unnamed: 0')
self.filedf.dropna(inplace=True)
#Calculate deal price,
self.filedf['r0mean'] = (self.filedf['r2mean'] - 0) * 0.3
self.filedf['r1mean'] = self.filedf['r0mean'] + (
self.filedf['r2mean'] - self.filedf['r0mean']) / 2
self.filedf['deal2'] = self.filedf['r2mean'] + (
self.filedf['r3mean'] - self.filedf['r2mean']) * 0.3
self.filedf['deal3'] = self.filedf['r3mean'] + (
self.filedf['r4mean'] - self.filedf['r3mean']) * 0.3
self.filedf['deal4'] = self.filedf['r4mean'] + (
self.filedf['r5mean'] - self.filedf['r4mean']) * 0.3
self.filedf['deal5'] = self.filedf['r5mean'] + (
self.filedf['r6mean'] - self.filedf['r5mean']) * 0.3
self.filedf['deal6'] = self.filedf['r6mean'] + (
self.filedf['r7mean'] - self.filedf['r6mean']) * 0.3
self.filedf['deal7'] = self.filedf['r7mean'] + (
self.filedf['r8mean'] - self.filedf['r7mean']) * 0.3
self.filedf['deal8'] = self.filedf['r8mean'] + (
self.filedf['r9mean'] - self.filedf['r8mean']) * 0.3
self.filedf['deal9'] = self.filedf['r9mean'] + (
self.filedf['r10mean'] - self.filedf['r9mean']) * 0.3
self.filedf['highest'] = self.filedf[[
'r0mean', 'r1mean', 'r2mean', 'r3mean', 'r4mean', 'r5mean',
'r6mean', 'r7mean', 'r8mean', 'r9mean'
]].max(axis=1)
self.filedf['lowest'] = self.filedf[[
'r0mean', 'r1mean', 'r2mean', 'r3mean', 'r4mean', 'r5mean',
'r6mean', 'r7mean', 'r8mean', 'r9mean'
]].min(axis=1)
self.filedf['profit'] = self.filedf['highest'] - self.filedf['lowest']
self.filedf['risk'] = self.filedf['profit'] / 2
self.filedf['maxname'] = self.filedf[[
'r0mean', 'r1mean', 'r2mean', 'r3mean', 'r4mean', 'r5mean',
'r6mean', 'r7mean', 'r8mean', 'r9mean'
]].idxmax(axis=1)
self.filedf['minname'] = self.filedf[[
'r0mean', 'r1mean', 'r2mean', 'r3mean', 'r4mean', 'r5mean',
'r6mean', 'r7mean', 'r8mean', 'r9mean'
]].idxmin(axis=1)
maxlist = list(map(itemgetter(1), self.filedf['maxname'].values))
self.filedf['maxtime'] = maxlist
self.filedf["maxtime"] = pd.to_numeric(self.filedf["maxtime"])
minlist = list(map(itemgetter(1), self.filedf['minname'].values))
self.filedf['mintime'] = minlist
self.filedf["mintime"] = pd.to_numeric(self.filedf["mintime"])
self.filedf[
'order_type'] = self.filedf['maxtime'] > self.filedf['mintime']
self.filedf['placetimebuy'] = self.filedf['mintime'] * 60
self.filedf['endtimebuy'] = self.filedf['maxtime'] * 60
self.filedf['placetimesell'] = self.filedf['maxtime'] * 60
self.filedf['endtimesell'] = self.filedf['mintime'] * 60
self.filedf.sort_values('profit', inplace=True, ascending=False)
self.filedf = self.filedf[self.filedf['total_hits'] >= 20]
self.filedf.sort_values('profit', inplace=True, ascending=False)
self.filedf = self.filedf.head(int(len(self.filedf) * 0.5))
temp = self.filedf
self.usabledf = temp
genes = temp.index.tolist()
pool_genes = []
for x in genes:
pool_genes.append(ast.literal_eval(x))
self.genes = pool_genes
df = pd.read_csv('df.csv')
_, self.binDict = qe.makeDf(df)
self.inverseGenes()