def Implement_RSI(self,
sec_id='AAPL',
qty=10,
period=14,
start_date="2018-10-30"):
# Get real time data from Interactive Broker
user1 = DataAPI()
df = user1.yahoo_pricing_data(sec_id, start_date=start_date)
# Check the signal
# -- sma --
BT_user1 = Backtesting(df)
df = BT_user1.RSI_single_stock(period=14)
print('signal: {}'.format(df.loc[df.shape[0] - 1, 'signal']))
# check the most recent signal
if df.loc[df.shape[0] - 1, 'signal'] == 'buy':
self.submit_buy_order(sec_id, qty)
self.Email_client.send_buy_notification(self.email_add, sec_id)
elif df.loc[df.shape[0] - 1, 'signal'] == 'sell':
self.submit_sell_order(sec_id, qty)
self.Email_client.send_sell_notification(self.email_add, sec_id)
def __init__(self, sec_id: str = 'AAPL'):
self.sec_id = sec_id
self.sf_data_client = DataAPI(which_API = 'SimFin')
self.yh_data_client = DataAPI(which_API = 'yahoo')
self.period = 'annual'
self.market = 'us'
self.IS_revenue = None
def get_resource(self, resource, params):
"""
:param resource: the twitter api resource, eg "followers/list"
:param params: resource params, eg { "screen_name": "foo", "cursor": "-1"}
:param data_field: what field actually contains the data results, eg data["users"]
:return: a tuple of next_cursor, data. next_cursor can be None.
"""
url = TwitterAPI.get_url(resource, params)
print url
data = self.get_response(url)
summary = None
if data:
other_params = data
if type(data) is list:
other_params = data[0]
if 'summarize_filename_prefix' in resource:
for field in resource['summarize_filename_prefix']:
keys = field.split("|")
value = DataAPI.find_key(other_params, keys)
if value:
params[field] = value
else:
resource['filename_fields'].remove(field)
file_path = TwitterAPI.get_filename(resource['url'], params, resource['filename_fields'])
TwitterAPI.save_resource(data, file_path)
summary = TwitterAPI.get_summary(data, params, resource)
return summary, data
def get_resource(self, resource, params):
"""
:param resource: the twitter api resource, eg "followers/list"
:param params: resource params, eg { "screen_name": "foo", "cursor": "-1"}
:param data_field: what field actually contains the data results, eg data["users"]
:return: a tuple of next_cursor, data. next_cursor can be None.
"""
url = TwitterAPI.get_url(resource, params)
print url
data = self.get_response(url)
summary = None
if data:
other_params = data
if type(data) is list:
other_params = data[0]
if 'summarize_filename_prefix' in resource:
for field in resource['summarize_filename_prefix']:
keys = field.split("|")
value = DataAPI.find_key(other_params, keys)
if value:
params[field] = value
else:
resource['filename_fields'].remove(field)
file_path = TwitterAPI.get_filename(resource['url'], params,
resource['filename_fields'])
TwitterAPI.save_resource(data, file_path)
summary = TwitterAPI.get_summary(data, params, resource)
return summary, data
class Analysis_report:
def __init__(self, sec_id: str = 'AAPL'):
self.sec_id = sec_id
self.sf_data_client = DataAPI(which_API = 'SimFin')
self.yh_data_client = DataAPI(which_API = 'yahoo')
self.period = 'annual'
self.market = 'us'
self.IS_revenue = None
def stock_report(self, File_path):
#// Retrive pricing data from yahoo finance
df_pricing = self.yh_data_client.yahoo_pricing_data(sec_id = self.sec_id)
#// Retrieve Income Statement
df_income = self.sf_data_client.SF_income_statement(sec_id = self.sec_id, period = self.period, market = self.market )
#// Retrieve Balance Sheet Data
df_bs = self.sf_data_client.SF_balance_sheet(sec_id = self.sec_id, period = self.period, market = self.market )
#// Retrieve Cash Flow Statement
df_cf = self.sf_data_client.SF_cashflow_statment(sec_id = self.sec_id, period = self.period, market = self.market )
#// Retrieve Ratios
df_ratios = self.sf_data_client.SF_Derived_Ratios(sec_id = self.sec_id, period = self.period, market= self.market )
#// Plot pricing chart
self.plotly_stock_price(df_pricing)
#// Balance Sheet
df_bs = df_bs.T
#// Balance Sheet commonsize
df_BS_common_size = self.balance_sheet_commonsize(df_bs.T)
#// Income Statement Common Size
df_IS_common_size = self.income_statment_common_size(df_income)
#// Cash flow statement
df_cf = df_cf.T
#// Ratio
df_ratios = self.ratios_yoy(df_ratios)
#// write data to an excel file
with pd.ExcelWriter(File_path) as writer:
df_bs.to_excel(writer, sheet_name = "Balance Sheet")
df_BS_common_size.to_excel(writer, sheet_name = "Balance Sheet Commonsize")
df_IS_common_size.to_excel(writer, sheet_name='Income Statement')
df_cf.to_excel(writer, sheet_name = "Cash Flow Statement")
df_ratios.to_excel(writer, sheet_name = "Ratios")
df_pricing.to_excel(writer, sheet_name = "Historical_price")
def plotly_stock_price(self, df):
fig = go.Figure(data=[go.Candlestick(x=df['date'],
open=df['open'],
high=df['high'],
low=df['low'],
close=df['close'])])
fig.layout.xaxis.type = 'category'
fig.show()
def income_statment_common_size(self, df):
dfT = df.T
Row_Revenue = dfT.index.get_loc('Revenue') + 1
Row_lis = dfT.index[Row_Revenue:]
self.IS_revenue = dfT.loc['Revenue']
df_temp = dfT.iloc[:Row_Revenue, :]
for i in Row_lis:
S_lis = self.IS_add_yoy_rows(dfT.loc[i])
df_temp = df_temp.append(dfT.loc[i].map(lambda x: '{:,} M'.format(x/1000000)), ignore_index = False)
df_temp = df_temp.append(S_lis, ignore_index = False)
return df_temp
def IS_add_yoy_rows(self, row_series):
#// Add yoy pct change, commonsize, yoy_commonsize rows into Income Stetement
row_index_name = row_series.name
YOY_name = "YOY_" + row_index_name
S_YOY_name = pd.Series(self.yoy_pct_change(row_series), name = YOY_name)
CommonSize_Name= "CommonSize_" + row_index_name
S_CommonSize_Name = pd.Series(row_series/self.IS_revenue, name = CommonSize_Name)
YOY_CommonSize_Name = "YOY_CommonSize_" + row_index_name
S_YOY_CommonSize_Name = pd.Series(self.yoy_pct_change(S_CommonSize_Name), name = YOY_CommonSize_Name)
S_lis = [S_YOY_name, S_CommonSize_Name, S_YOY_CommonSize_Name]
return S_lis
def yoy_pct_change(self, col_series):
#// replace pct_change function as it is not accurate when numbers are negative
S_pct_change = (col_series - col_series.shift(1))/col_series.shift(1).map(lambda x: x if x >= 0 else -x)
S_pct_change = S_pct_change.map(lambda x: '{:.2%}'.format(x))
return S_pct_change
def balance_sheet_yoy(self, df):
dfT = df.T
Ros_shares = dfT.index.get_loc('Shares (Diluted)') + 1
Row_lis = dfT.index[Ros_shares:]
df_temp = dfT.iloc[:Ros_shares, :]
for i in Row_lis:
S_yoy_col = self.BS_add_yoy_rows(dfT.loc[i])
df_temp = df_temp.append(dfT.loc[i].map(lambda x: '{:,} M'.format(x/1000000)), ignore_index = False)
df_temp = df_temp.append(S_yoy_col, ignore_index = False)
return df_temp
def BS_add_yoy_rows(self, row_series):
#// Add yoy pct change, commonsize, yoy_commonsize rows into Income Stetement
row_index_name = row_series.name
YOY_name = "YOY_" + row_index_name
S_YOY_name = pd.Series(self.yoy_pct_change(row_series), name = YOY_name)
return S_YOY_name
def yoy_pct_change(self, col_series):
#// replace pct_change function as it is not accurate when numbers are negative
col_shift_1 = col_series.shift(1).map(lambda x: x + 0.0001 if x >= 0 else -x)
S_pct_change = (col_series - col_series.shift(1))/col_shift_1
S_pct_change = S_pct_change.map(lambda x: '{:.2%}'.format(x))
return S_pct_change
def balance_sheet_commonsize(self, df):
#// transpose to date as columns
df = df.T
#// categorize index names into asset and liabilities
lis_asset = ['Cash, Cash Equivalents & Short Term Investments', 'Accounts & Notes Receivable', 'Inventories',
'Property, Plant & Equipment, Net', 'Long Term Investments & Receivables', 'Other Long Term Assets',
'Total Noncurrent Assets']
total_asset = 'Total Assets'
lis_liability = ['Payables & Accruals','Short Term Debt', 'Total Current Liabilities',
'Long Term Debt','Total Noncurrent Liabilities']
total_liability = 'Total Liabilities'
lis_equity = ['Share Capital & Additional Paid-In Capital', 'Treasury Stock','Retained Earnings']
total_equity = 'Total Equity'
total_liabilityNequity = 'Total Liabilities & Equity'
df.loc[lis_asset] = df.loc[lis_asset]/df.loc[total_asset]
df.loc[lis_liability] = df.loc[lis_liability]/df.loc[total_liabilityNequity]
df.loc[lis_equity] = df.loc[lis_equity]/df.loc[total_liabilityNequity]
lis_a = lis_asset + lis_liability + lis_equity
lis_b = [total_asset, total_liability, total_equity, total_liabilityNequity, 'Total Current Assets']
df.loc[lis_a] = df.loc[lis_a].applymap(lambda x: '{:.2%}'.format(x))
df.loc[lis_b] = df.loc[lis_b].applymap(lambda x: '{:,} M'.format(x/1000000))
return df
def ratios_yoy(self, df):
df = df.T
lis = ['EBITDA', 'Total Debt', 'Free Cash Flow',
'Gross Profit Margin', 'Operating Margin', 'Net Profit Margin',
'Return on Equity', 'Return on Assets', 'Free Cash Flow to Net Income',
'Current Ratio', 'Liabilities to Equity Ratio', 'Debt Ratio',
'Earnings Per Share, Basic', 'Earnings Per Share, Diluted',
'Sales Per Share', 'Equity Per Share', 'Free Cash Flow Per Share',
'Dividends Per Share']
Row_Ebida = df.index.get_loc('EBITDA')
df_temp = df.iloc[:Row_Ebida, :]
for i in lis:
row_index_name = df.loc[i].name
YOY_name = "YOY_" + row_index_name
S_YOY_name = pd.Series(self.yoy_pct_change(df.loc[i]), name = YOY_name)
df_temp = df_temp.append(df.loc[i])
df_temp = df_temp.append(S_YOY_name, ignore_index = False)
lis_dollar_Form = ['EBITDA', 'Total Debt', 'Free Cash Flow']
lis_pct_form = ['Gross Profit Margin', 'Operating Margin', 'Net Profit Margin',
'Return on Equity', 'Return on Assets', 'Free Cash Flow to Net Income',
'Current Ratio', 'Liabilities to Equity Ratio', 'Debt Ratio']
df_temp.loc[lis_pct_form] = df_temp.loc[lis_pct_form].applymap(lambda x: '{:.2%}'.format(x))
df_temp.loc[lis_dollar_Form] = df_temp.loc[lis_dollar_Form].applymap(lambda x: '{:,} M'.format(x/1000000))
return df_temp
from data import DataAPI
from backtesting import Backtesting
from Visualization import Visualization
import pandas as pd
import matplotlib.pyplot as plt
# user_yf = DataAPI('yahoo')
# df = user_yf.yahoo_pricing_data()
# print(df)
user1 = DataAPI("SimFin")
# df_income = user1.SF_income_statement()
# print('Income statement')
# print(df_income)
# df_bs = user1.SF_balance_sheet()
# print('balancesheet statement')
# print(df_bs)
# df_cf = user1.SF_cashflow_statment()
# print('cashflow statement')
# print(df_cf)
#// test sma_backtesting single
# df = user1.SF_pricing_data(start_date='2016-01-01')
# # df = user1.SF_daily_pricing_data_from_local()
from data import DataAPI
from backtesting import Backtesting
from Visualization import Visualization
import pandas as pd
import matplotlib.pyplot as plt
user1 = DataAPI("yahoo")
df = user1.yahoo_pricing_data('AAPL', start_date= "2015-10-30")
# -- sma --
# bt1 = Backtesting(df)
# df1 = bt1.backtesting_sma_single_stock(sma_short=5, sma_long=30)
# v1 = Visualize(df1)
# v1.sma_plot_result()
# -- rsi --
# bt1 = Backtesting(df)
# df1 = bt1.RSI_single_stock(period= 28)
# v1 = Visualize(df1)
# v1.sma_plot_result()
# path = r'D:\K\HU\HU - Courses\CISC 695 Research Methodology and Writing\Assignments\sample_data\rsi_aapl_data.csv'
# df1.to_csv(path)
#// test sma_backtesting bulk
def predict_api(data: DataAPI):
# Extract data in correct order
df_data = pd.DataFrame(data.dict(), index=[0])
to_predict = prepare_data_api(df_data)
response = xgb.predict(to_predict)
return {'prediction': float(response[0])}
def get_data(self):
df = DataAPI.yahoo_pricing_data('AAPL', start_date='2019-01-01', end_date= '2020-10-30')
return df
def test_SF_daily_pricing_data(self):
usr = DataAPI(which_API='SimFin', token ='free')
result = usr.SF_pricing_data('AAPL', period='Daily', market = 'us',
start_date='2018-01-01', end_date= '2020-10-30')
self.assertIsInstance(result, pd.DataFrame)
def test_yahoo_pricing_data(self):
usr = DataAPI(which_API='yahoo')
result = usr.yahoo_pricing_data('AAPL', start_date='2018-01-01', end_date= '2020-10-30')
self.assertIsInstance(result, pd.DataFrame)
def test_SF_cashflow_statment(self):
usr = DataAPI(which_API='SimFin', token ='free')
result = usr.SF_cashflow_statment('AAPL', period='quarterly', market = 'us')
self.assertIsInstance(result, pd.DataFrame)
from data import DataAPI
from backtesting import Backtesting
from Visualization import Visualize
import pandas as pd
import matplotlib.pyplot as plt
user1 = DataAPI("IEX", IEX_token)
# path_aapl2y = r'D:\K\HU\HU - Courses\CISC 695 Research Methodology and Writing\Assignments\sample_data\aapl2y.csv'
# df = pd.read_csv(path_aapl2y)
# df = pd.read_csv(path_aapl2y)
df = user1.IEX_daily_pricing_data('AAPL', '2y')
bt1 = Backtesting(df)
df1 = bt1.RSI_single_stock(notification=False)
v1 = Visualize(df1)
v1.sma_plot_result()
# df2 = bt1.sma_backteting_bulk()
# print(df2)