def compilereturnsdataframe(self
, showresults = 0):
#'^GSmr ^OEX ^VIX ^OEX ^MID ^RUT ^DJI
# if self.Period == 'daily':
# import pullreturnsdaily as pr
# else:
# import pullreturnsmonthly as pr
import pullreturnsbyperiod as pr
dict_of_df_pricehistory = {}
dict_of_df_returns = {}
# Build the annualized returns series
mysymbolslist = self.SymbolsList # list_of_symbols #['^GSPC','^DJI','^MID','^OEX','AAPL','LEO']
ser_annual = pd.Series()
for symbol in mysymbolslist:
o = pr.perform(symbol,self.StartDateString,self.Period)
#print o.ReturnsDataframe[['e_'+ self.PctChangeOrLogReturn]]
#df_pricehistorydataframe = o.PriceHistoryDataframe
#dfcompilereturnsdataframe = o.ReturnsDataframe
dict_of_df_pricehistory[symbol] = o.PriceHistoryDataframe
dict_of_df_returns[symbol] = o.ReturnsDataframe
annualizedreturn = o.annualizedreturn()
ser_annual = ser_annual.set_value(symbol, annualizedreturn)
self.AnnualizedReturnsSeries = ser_annual
#mysymbolslist = self.SymbolsList # list_of_symbols #['^GSPC','^DJI','^MID','^OEX','AAPL','LEO']
#ser_annual = pd.Series()
# for symbol in mysymbolslist:
# df_pricehistory = dict_of_df_pricehistory[symbol]
# df_returns = dict_of_df_returns[symbol]
# o = pr.perform(symbol,self.StartDateString)
# df = o.ReturnsDataframe
# dict_of_dfs[symbol] = df
index = ['X']
columns = ['A','B', 'C']
df_largestofreturns = pd.DataFrame(index=index, columns=columns)
df_largestofreturns = df_largestofreturns.fillna(0) # with 0s rather than NaNs
#print df_largestofreturns
#while len(dict_of_dfs_bysize) < len(dict_of_dfs):
keyoflargestreturnsdf = ''
for k,v in dict_of_df_returns.items():
if len(v) > len(df_largestofreturns):
df_largestofreturns = v
keyoflargestreturnsdf = k
#break
#wwwwww
# print df_largestofreturns
# print keyoflargestreturnsdf
df_alignedreturns = df_largestofreturns[['e_'+ self.PctChangeOrLogReturn]]
#df_alignedreturns = df_alignedreturns.set_index('b_periodend')
df_alignedreturns.columns = [keyoflargestreturnsdf]
df_alignedreturns.sort_index
#print df_alignedreturns
#print df_largestofreturns
#
# if passed == 0:
# df_alignedreturns = v[['b_periodend','e_pctchange']]
# df_alignedreturns = df_alignedreturns.set_index('b_periodend')
# df_alignedreturns.columns = [k]
# df_alignedreturns.sort_index
# #sLength = len(df_alignedreturns[k])
# #originalid = k
#
# else:
for k,v in dict_of_df_returns.items():
if not k == keyoflargestreturnsdf:
df_new = v[['e_' + self.PctChangeOrLogReturn]]
#df_new = df_new.set_index('b_periodend')
df_new.columns = [k]
df_new.sort_index
#print df_new
#df_alignedreturns[k] = df_new.loc[k].shape[0]
#print df_new
#df_alignedreturns[k] = pd.Series(df_new, index=df_alignedreturns.index)
#df_alignedreturns[k] = df_alignedreturns[originalid].map(lambda x: df_new[k])
df_alignedreturns[k] = df_new[k]
if showresults == 1:
print '----------------------------------------------------'
print ' returns'
print '----------------------------------------------------'
print df_alignedreturns
#print df_alignedreturns
#self.MonthlyReturnsDataframe = df_alignedreturns
return df_alignedreturns
def compilehistoricaldataframes(self
, showresults = 0):
#'^GSmr ^OEX ^VIX ^OEX ^MID ^RUT ^DJI
# if self.Period == 'daily':
# import pullreturnsdaily as pr
# else:
# import pullreturnsmonthly as pr
import pullreturnsbyperiod as pr
dict_of_df_pricehistory = {}
dict_of_df_returns = {}
# Build the annualized returns series
mysymbolslist = self.SymbolsList # list_of_symbols #['^GSPC','^DJI','^MID','^OEX','AAPL','LEO']
ser_annual = pd.Series()
for symbol in mysymbolslist:
o = pr.perform(symbol,self.StartDateString,self.Period)
dict_of_df_pricehistory[symbol] = o.PriceHistoryDataframe
dict_of_df_returns[symbol] = o.ReturnsDataframe
annualizedreturn = o.annualizedreturn()
ser_annual = ser_annual.set_value(symbol, annualizedreturn)
self.AnnualizedReturnsSeries = ser_annual
index = ['X']
columns = ['A','B', 'C']
# ----------------------------------------------------------------------
df_largestofreturns = pd.DataFrame(index=index, columns=columns)
df_largestofreturns = df_largestofreturns.fillna(0) # with 0s rather than NaNs
keyoflargestreturnsdf = ''
for k,v in dict_of_df_returns.items():
if len(v) > len(df_largestofreturns):
df_largestofreturns = v
keyoflargestreturnsdf = k
df_alignedreturns = df_largestofreturns[['e_'+ self.PctChangeOrLogReturn]]
df_alignedreturns.columns = [keyoflargestreturnsdf]
df_alignedreturns.sort_index
for k,v in dict_of_df_returns.items():
if not k == keyoflargestreturnsdf:
df_new = v[['e_' + self.PctChangeOrLogReturn]]
df_new.columns = [k]
df_new.sort_index
df_alignedreturns[k] = df_new[k]
self.ReturnsDataframe = df_alignedreturns
# ----------------------------------------------------------------------
df_largestofpricehistory = pd.DataFrame(index=index, columns=columns)
df_largestofpricehistory = df_largestofpricehistory.fillna(0) # with 0s rather than NaNs
keyoflargestpricehistorydf = ''
for k,v in dict_of_df_pricehistory.items():
if len(v) > len(df_largestofpricehistory):
df_largestofpricehistory = v
keyoflargestpricehistorydf = k
df_alignedpricehistory = df_largestofpricehistory[['e_adjclose']]
df_alignedpricehistory.columns = [keyoflargestpricehistorydf]
df_alignedpricehistory.sort_index
for k,v in dict_of_df_pricehistory.items():
if not k == keyoflargestpricehistorydf:
df_new = v[['e_adjclose']]
df_new.columns = [k]
df_new.sort_index
df_alignedpricehistory[k] = df_new[k]
self.PriceHistoryDataframe = df_alignedpricehistory
def pricehistorydataframe(self
, period = 'monthly' # or daily
, showresults = 0):
#'^GSmr ^OEX ^VIX ^OEX ^MID ^RUT ^DJI
import pullreturnsbyperiod as pr
# if self.Period == 'daily':
# import pullreturnsdaily as pr
# else:
# import pullreturnsmonthly as pr
dict_of_dfs = {}
print 'xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx'
# Build the annualized returns series
mysymbolslist = self.SymbolsList # list_of_symbols #['^GSPC','^DJI','^MID','^OEX','AAPL','LEO']
#ser_annual = pd.Series()
for symbol in mysymbolslist:
o = pr.perform(symbol,self.StartDateString,period)
df = o.PriceHistoryDataframe
dict_of_dfs[symbol] = df
#annualizedreturn = o.annualizedreturn()
#ser_annual = ser_annual.set_value(symbol, annualizedreturn)
# if not showresults == 0:
# print '----- ser_annual -----'
# print ser_annual
#print 'returns dataframe ready'
#self.AnnualizedReturnsSeries = ser_annual
# mysymbolslist = self.SymbolsList # list_of_symbols #['^GSPC','^DJI','^MID','^OEX','AAPL','LEO']
# #ser_annual = pd.Series()
# for symbol in mysymbolslist:
# o = pr.perform(symbol,self.StartDateString)
# df = o.ReturnsDataframe
# dict_of_dfs[symbol] = df
index = ['X']
columns = ['A','B', 'C']
df_largestofreturns = pd.DataFrame(index=index, columns=columns)
df_largestofreturns = df_largestofreturns.fillna(0) # with 0s rather than NaNs
#print df_largestofreturns
#while len(dict_of_dfs_bysize) < len(dict_of_dfs):
keyoflargestreturnsdf = ''
for k,v in dict_of_dfs.items():
if len(v) > len(df_largestofreturns):
df_largestofreturns = v
keyoflargestreturnsdf = k
break
print '^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^', keyoflargestreturnsdf
df_alignedreturns = df_largestofreturns[['e_adjclose']]
#print df_alignedreturns
df_alignedreturns.columns = [keyoflargestreturnsdf]
#df_alignedreturns = df_alignedreturns.set_index('b_periodend')
#print df_alignedreturns
df_alignedreturns.sort_index
#print df_alignedreturns
# if passed == 0:
# df_alignedreturns = v[['b_periodend','e_pctchange']]
# df_alignedreturns = df_alignedreturns.set_index('b_periodend')
# df_alignedreturns.columns = [k]
# df_alignedreturns.sort_index
# #sLength = len(df_alignedreturns[k])
# #originalid = k
#
# else:
for k,v in dict_of_dfs.items():
if not k == keyoflargestreturnsdf:
df_new = v[['e_adjclose']]
#print '----------------v--------------------'
#print v
#print '----------------df_new--------------------'
#print df_new
#df_new = df_new.set_index('b_periodend')
df_new.columns = [k]
#print df_new
df_new.sort_index
#print 'made it here'
#print df_new
#df_alignedreturns[k] = df_new.loc[k].shape[0]
#print df_new
#df_alignedreturns[k] = pd.Series(df_new, index=df_alignedreturns.index)
#df_alignedreturns[k] = df_alignedreturns[originalid].map(lambda x: df_new[k])
df_alignedreturns[k] = df_new#[k]
if showresults == 1:
print '----------------------------------------------------'
print ' stockhistory'
print '----------------------------------------------------'
print '------------------------------df_alignedreturns -------------------------'
print df_alignedreturns
#self.MonthlyReturnsDataframe = df_alignedreturns
return df_alignedreturns