Skip to content

Commit

Permalink
Update famafrench.py
Browse files Browse the repository at this point in the history
  • Loading branch information
@christianjauregui
christianjauregui committed May 2, 2020
1 parent ff524ca commit 5448a97
Showing 1 changed file with 25 additions and 16 deletions.
41 changes: 25 additions & 16 deletions famafrench/famafrench.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -20,6 +20,8 @@
__all__ = ["Error", "FamaFrench"]

# Standard Imports
import os
import errno
import pickle
import copy
import sqlalchemy
Expand Down Expand Up @@ -84,7 +86,9 @@ class FamaFrench:
`wrds-cloud <https://wrds-www.wharton.upenn.edu/>`_.
Attributes
-----------
-----------
pickled_dir : str
Absolute path directory where pickled datasets will be saved to.
runQuery: bool
Flag for choosing whether to query datafiles from `wrds-cloud <https://wrds-www.wharton.upenn.edu/>`_ or import locally-pickled files.
freqType: str
Expand Down Expand Up @@ -137,16 +141,14 @@ class FamaFrench:
Flag for choosing whether to run any estimation procedures for the first time/re-estimate procedures
by querying `wrds-cloud <https://wrds-www.wharton.upenn.edu/>`_ datafiles
`or` import locally-pickled files storing existing estimates.
pickled_directory : str, default `\'./famafrench/pickled_db/\'`
Local directory extention where pickled datasets will be saved to.
runFactorReg : bool, default True
Flag for initializing the estimation of market beta's and/or rolling residual variances required
for sorting portfolios on ``BETA`` and/or ``RESVAR``.
This boolean is set internally - no need for user to specify it externally.
"""
# Constructor for the 'FamaFrench' object parent class
def __init__(self, runQuery, freqType, sortCharacsId, factorsId, mainCharacsId=None, runEstimation=False, pickled_directory='./famafrench/pickled_db/'):
def __init__(self, pickled_dir, runQuery, freqType, sortCharacsId, factorsId, mainCharacsId=None, runEstimation=False):
"""
Constructor for the 'FamaFrench' object parent class
"""
Expand All @@ -173,8 +175,15 @@ def __init__(self, runQuery, freqType, sortCharacsId, factorsId, mainCharacsId=N
self.runQuery = runQuery
self.runEstimation = runEstimation

# Pickled datasets are saved in a local directory w/ the given name.
self.pickled_directory = pickled_directory
# Pickled datasets are saved in the directory w/ a specific absolute path:
for subdir in ['', 'daily/', 'monthly/', 'weekly/', 'quarterly/', 'annual/']:
if not os.path.exists(os.path.dirname(pickled_dir + subdir)):
try:
os.makedirs(os.path.dirname(pickled_dir + subdir))
except OSError as exc: # Guard against race condition
if exc.errno != errno.EEXIST:
raise
self.pickled_dir = pickled_dir


@utils.lru_cached_method(maxsize=32)
Expand Down Expand Up @@ -307,7 +316,7 @@ def get_comp_sqlQuery(wrds_update, start_date, end_date):
return comp_sqlQuery

# Load local file if it exists (and dates can be found), else query from wrds-cloud.
comp_file = Path(self.pickled_directory + 'annual/comp_annual.pickle')
comp_file = Path(self.pickled_dir + 'annual/comp_annual.pickle')
if self.runQuery is False and comp_file.is_file():
file_to_pickle = open(comp_file, 'rb')
dfcomp = pickle.load(file_to_pickle)
Expand Down Expand Up @@ -612,7 +621,7 @@ def get_crsp_sqlQuery(wrds_update, freqTypeFullstr, start_date, end_date):
return crsp_sqlQuery

# Load local file if it exists (and dates can be found), else query from wrds-cloud.
crsp_file = Path(self.pickled_directory + freqTypeFull + '/crsp_' + freqTypeFull + '.pickle')
crsp_file = Path(self.pickled_dir + freqTypeFull + '/crsp_' + freqTypeFull + '.pickle')
if self.runQuery is False and crsp_file.is_file():
file_to_pickle = open(crsp_file, 'rb')
dfcrsp = pickle.load(file_to_pickle)
Expand Down Expand Up @@ -782,7 +791,7 @@ def get_crspdlret_sqlQuery(wrds_update, freqType):
return crspdlret_sqlQuery

# Load local file if it exists (and dates can be found), else query from wrds-cloud.
crspdlret_file = Path(self.pickled_directory + freqTypeFull + '/crspdelist_' + freqTypeFull + '.pickle')
crspdlret_file = Path(self.pickled_dir + freqTypeFull + '/crspdelist_' + freqTypeFull + '.pickle')
if self.runQuery is False and crspdlret_file.is_file():
file_to_pickle = open(crspdlret_file, 'rb')
dfcrspdlret = pickle.load(file_to_pickle)
Expand Down Expand Up @@ -959,7 +968,7 @@ def get_rf1m_sqlQuery(freqType, start_date, end_date):
return rf1m_sqlQuery

# Load local file if it exists (and dates can be found), else query from wrds-cloud.
rf1m_file = Path(self.pickled_directory + freqTypeFulldir + '/rf1m_' + freqTypeFulldir + '.pickle')
rf1m_file = Path(self.pickled_dir + freqTypeFulldir + '/rf1m_' + freqTypeFulldir + '.pickle')
if self.runQuery is False and rf1m_file.is_file():
file_to_pickle = open(rf1m_file, 'rb')
rf1m = pickle.load(file_to_pickle)
Expand Down Expand Up @@ -1179,7 +1188,7 @@ def get_crsprollvar_sqlQuery(wrds_update, freqType, roll_win, min_per, start_dat
return crsprollvar_sqlQuery

# Load local file if it exists (and dates can be found), else query from wrds-cloud.
crsprollvar_file = Path(self.pickled_directory + freqTypeFull + '/crsprollvar_' + freqTypeFull + '.pickle')
crsprollvar_file = Path(self.pickled_dir + freqTypeFull + '/crsprollvar_' + freqTypeFull + '.pickle')
if self.runQuery is False and crsprollvar_file.is_file():
file_to_pickle = open(crsprollvar_file, 'rb')
dfcrsprollvar = pickle.load(file_to_pickle)
Expand Down Expand Up @@ -1498,7 +1507,7 @@ def getFactorRegResults(self, ffmodel, boolRVar, scholeswilliams, roll_window, m
# Need to query AT A MINIMUM (minus) -'roll_window' days before 'dt_start' so we have a non-missing estimation for 'dt_start':
dt_start = (dt_start - Day(roll_window)).date()

crsp_ret_file = Path(self.pickled_directory + freqTypeFull + '/crspretFactorReg_' + freqTypeFull + '.pickle')
crsp_ret_file = Path(self.pickled_dir + freqTypeFull + '/crspretFactorReg_' + freqTypeFull + '.pickle')
if self.runEstimation is False and crsp_ret_file.is_file():
file_to_pickle = open(crsp_ret_file, 'rb')
dfcrsp_ret = pickle.load(file_to_pickle)
Expand Down Expand Up @@ -1573,7 +1582,7 @@ def getFactorRegResults(self, ffmodel, boolRVar, scholeswilliams, roll_window, m
else:
raise TypeError('\'ffmodel\' should be the Market model (i.e. CAPM) or the Fama-French 3 model:\n use \'capm\' or \'ff3\'.')

ff_file = Path(self.pickled_directory + freqTypeFull + '/' + ffmodel + 'factors_' + freqTypeFull + '.pickle')
ff_file = Path(self.pickled_dir + freqTypeFull + '/' + ffmodel + 'factors_' + freqTypeFull + '.pickle')
if self.runEstimation is False and ff_file.is_file():
file_to_pickle = open(ff_file, 'rb')
dfportSort_tableList = pickle.load(file_to_pickle)
Expand Down Expand Up @@ -2080,7 +2089,7 @@ class methods :meth:`famafrench.FamaFrench.mergeCCM` and :meth:`famafrench.FamaF
else:
freqTypeFull = 'monthly'

resvar_file = Path(self.pickled_directory + freqTypeFull + '/crspff3resvar_' + freqTypeFull + '.pickle')
resvar_file = Path(self.pickled_dir + freqTypeFull + '/crspff3resvar_' + freqTypeFull + '.pickle')
if self.runEstimation is False and resvar_file.is_file():
file_to_pickle = open(resvar_file, 'rb')
dfresvarff3 = pickle.load(file_to_pickle)
Expand Down Expand Up @@ -2174,7 +2183,7 @@ class methods :meth:`famafrench.FamaFrench.mergeCCM` and :meth:`famafrench.FamaF
else:
freqTypeFull, freqtmp = 'monthly', 'M'

beta_file = Path(self.pickled_directory + freqTypeFull + '/beta_' + freqTypeFull + '.pickle')
beta_file = Path(self.pickled_dir + freqTypeFull + '/beta_' + freqTypeFull + '.pickle')
if self.runEstimation is False and beta_file.is_file():
file_to_pickle = open(beta_file, 'rb')
dfbeta = pickle.load(file_to_pickle)
Expand Down Expand Up @@ -2393,7 +2402,7 @@ class method :meth:`famafrench.FamaFrench.getNyseThresholdsAndRet`.
"""

# Load local file if it exists, else query from wrds-cloud.
ccm_file = Path(self.pickled_directory + 'ccm.pickle')
ccm_file = Path(self.pickled_dir + 'ccm.pickle')
if self.runQuery is False and ccm_file.is_file():
file_to_pickle = open(ccm_file, 'rb')
dfccm = pickle.load(file_to_pickle)
Expand Down

0 comments on commit 5448a97

Please sign in to comment.