def get_last_trading_day_quotes(symbol):
today = datetime.today()
i = 1
yesterday = today - timedelta(i)
df = dr(symbol, 'yahoo', yesterday, today)
while df.empty:
i = i + 1
yesterday = today - timedelta(i)
df = dr(symbol, 'yahoo', yesterday, today)
return df
def __init__(self, ticker, json):
df = dr(ticker, 'yahoo', dt.datetime(1980, 01, 01))
pnl = pd.Series(np.zeros(len(df)), index=df.index)
for i in xrange(1, len(pnl)):
pnl[pnl.index[i]] = df['Adj Close'][df.index[i]] - df['Adj Close'][df.index[i - 1]]
df['PnL'] = pnl
price_list = []
pnl_list = []
if json:
newline = '\n'
else:
newline = '\\n'
for date, row in df.T.iteritems():
price_list.append(date.strftime('%Y-%m-%d') + ',' + str(row['Adj Close']) + newline)
pnl_list.append(date.strftime('%Y-%m-%d') + ',' + str(row['PnL']) + newline)
price_csv = 'Date,Price' + newline + ''.join(price_list)
pnl_csv = 'Date,PnL' + newline + ''.join(pnl_list)
# cache values using object attributes
self.ticker = ticker
self.dataframe = df
self.price_csv = price_csv
self.pnl_csv = pnl_csv
self.json = json
import numpy as np
import matplotlib.pyplot as plt
import pandas as pan
from pandas.io.data import DataReader as dr
import datetime
import hp_filter as hp
import Bandpass_Filter as bpf
from prettytable import PrettyTable as pt
# CPI: CPIAUCSL
# GDP: GDPC1
# consumption: PCECC96
# investment: GCEC96
gdp = np.asarray(
dr('GDPC1', 'fred', start = datetime.datetime(1947,1,1))['VALUE'])
cpi = np.asarray(
dr('CPIAUCSL', 'fred', start = datetime.datetime(1947,1,1))['VALUE'])
cons = np.asarray(
dr('PCECC96', 'fred', start = datetime.datetime(1947,1,1))['VALUE'])
inv = np.asarray(
dr('GCEC96', 'fred', start = datetime.datetime(1947,1,1))['VALUE'])
mask = np.arange(1,cpi.size, 3)
cpi = cpi[mask]
def problem_5():
"""
This function uses pyplot.psd to plot the spectrum for each of the time
series generated above.
import DSGE
import MonteCarloSimulation as mcs
import NumDeriv as nd
import SolveSS as sols
import UhligSolve as lig
import pandas as pan
from pandas.io.data import DataReader as dr
import datetime
##-------------------- Step 1. Get data and filter it.------------------------##
# We will use Wage data: ECIWAG
# Also business labor: PRS84006173
# and Consumption of fixed capital: COFC
start_date = datetime.datetime(2000, 1, 1)
wage_Data= np.asarray(dr('ECIWAG', 'fred', start = start_date)['VALUE'])
labor_Data = np.asarray(dr('PRS84006173', 'fred', start = start_date)['VALUE'])
consump_Data = np.asarray(dr('COFC', 'fred', start = start_date)['VALUE'])
filt_wage = hp.hp_filter(wage_Data)[1]
filt_labor = hp.hp_filter(labor_Data)[1]
filt_consump = hp.hp_filter(consump_Data)[1]
moments = np.array([np.mean(filt_consump), np.mean(filt_labor),
np.mean(filt_wage) ,np.var(filt_consump)])
##------------------------Step 2. Find steady state --------------------------##
def get_current_ss(beta):
"""
This calls SolveSS.py to get the steady state for the particular parameter
vector beta.
__author__ = 'Davidws'
import pandas as pd
from pandas.io.data import DataReader as dr
from datetime import datetime
import numpy as np
from numpy.linalg import inv
symbols = ['ORCL', 'MSFT', 'AAPL'] #Define the tickers we want in our project
start_date = '2010-01-01' # Give the start of the date range
d = {} # Create an empty dictionary to hold them
for ticker in symbols:
d[ticker] = dr(
ticker, 'yahoo', start_date
) # Iterate over the tickers and get the stock data into our dictionary
pan = pd.Panel(
d) # Create a Panel of the Data - this will allow us to Melt / Cast
close = pan.minor_xs(
'Adj Close'
) # Cast on the Adj Close - We could use any of the available columns here
# Careful about the tickers - some stocks are not this old and wont have data then we need to determine what to do about the NAN as we calculate the covariance.
count_tickers = symbols.__len__() # Get the Number of Stock Tickers we want
exp_return = close.mean(0) # Calculate the E(R) for the stocks
sd_stock = close.std # Calculate the Std.Dev for the stocks
cov_stock = close.cov() # Calculate the Covariance Matrix for the stocks
# Uncomment below if you want to write this to a file - add your own path
# path = ('C:/...)
# close.to_csv(path + 'close.csv', sep=',')
inv_cov = inv(
__author__ = 'Davidws'
import pandas as pd
from pandas.io.data import DataReader as dr
from datetime import datetime
import numpy as np
from numpy.linalg import inv
symbols = ['ORCL','MSFT','AAPL'] #Define the tickers we want in our project
start_date = '2010-01-01' # Give the start of the date range
d = {} # Create an empty dictionary to hold them
for ticker in symbols:
d[ticker] = dr(ticker, 'yahoo',start_date) # Iterate over the tickers and get the stock data into our dictionary
pan = pd.Panel(d) # Create a Panel of the Data - this will allow us to Melt / Cast
close = pan.minor_xs('Adj Close') # Cast on the Adj Close - We could use any of the available columns here
# Careful about the tickers - some stocks are not this old and wont have data then we need to determine what to do about the NAN as we calculate the covariance.
count_tickers = symbols.__len__() # Get the Number of Stock Tickers we want
exp_return = close.mean(0) # Calculate the E(R) for the stocks
sd_stock = close.std # Calculate the Std.Dev for the stocks
cov_stock = close.cov() # Calculate the Covariance Matrix for the stocks
# Uncomment below if you want to write this to a file - add your own path
# path = ('C:/...)
# close.to_csv(path + 'close.csv', sep=',')
inv_cov = inv(cov_stock) # Calculates the inverse of the covariance matrix - This may be going too far for our solution so far.