def test_fetch(self):
config_wrapper = loc_utils.use_test_configuration()
econ_platform_core.PlatformConfiguration = config_wrapper
econ_platform_core.init_package()
loc_utils.delete_data_file('TEST_TEST1.txt')
ser = econ_platform_core.fetch('TEST@TEST1', database='TEXT')
targ = get_test_series('TEST1')
self.assertTrue(targ.equals(ser))
self.assertTrue(
os.path.exists(
os.path.join(os.path.dirname(__file__), 'data',
'TEST_TEST1.txt')))
# fetch again, will be from file.
ser = econ_platform_core.fetch('TEST@TEST1', database='TEXT')
self.assertTrue(targ.equals(ser))
def test_user_function(self):
config_wrapper = loc_utils.use_test_configuration()
econ_platform_core.PlatformConfiguration = config_wrapper
econ_platform_core.init_package()
User_provider = econ_platform_core.Providers.UserProvider
User_provider.FunctionMapper['testfn'] = user_function
ser = econ_platform_core.fetch('U@testfn(0)', database='TEXT')
self.assertEqual(ser.values, [0])
def main():
"""
Routine used in testing. Not really a good candidate for test-driven development...
:return:
"""
econ_platform_core.configuration.print_configuration()
econ_platform_core.start_log()
obj = ProviderCansim_Csv()
# obj.ParseUnzipped('10100002')
ser = econ_platform_core.fetch('CCSV@10100002|v86822808')
def us_inflation(series_meta):
"""
:param series_meta: econ_platform_core.SeriesMetadata
:return:
"""
if not str(series_meta.ticker_query) == 'US_CPI_INFLATION':
raise ValueError('Wrong series!')
cpi_index = fetch('F@CPIAUCSL') # All -items, SA
inf = cpi_index.pct_change(12)
series_meta.series_name = '12-month Change in U.S. CPI Inflation'
series_meta.series_description = 'Calculated series'
return inf
def inflation_fn(series_meta, fn_args):
"""
Inflation dispersion in the U.S., based on major groups
food and beverages, housing, apparel, transportation, medical care, recreation, education and communication, and other goods and services
fn_args must be of the form N, return_type
N = # of months to do the analysis on (annualised inflation rate over N months).
return_type = "headline", "min", "max" (case insensitive, do not include quotes).
Either returns the headline inflation rate, or the min or max of the major groups.
:param series_meta: econ_platform_core.SeriesMetadata
:param fn_args': str
:return:
"""
try:
N_months, return_type = fn_args.split(',')
except:
raise ValueError(
'Dispersion function has two arguments: N, return_type')
# Remove any quotes from people who don't follow directions...
return_type = return_type.replace("'", '')
return_type = return_type.replace('"', '')
return_type = return_type.strip()
return_type = return_type.upper()
try:
N_months = int(N_months)
except:
raise ValueError('N_months must be an integer')
if return_type not in ('HEADLINE', 'MIN', 'MAX'):
raise ValueError(
f'Invalid return type {return_type}, must be one of HEADLINE, MIN, MAX'
)
if return_type == 'HEADLINE':
cpi_index = fetch('F@CPIAUCSL') # All -items, SA
pct_chg = cpi_index.pct_change(N_months)
factor = 1. + pct_chg
annualised = pow(factor, 12. / float(N_months))
annualised = 100. * (annualised - 1.)
annualised.name = str(series_meta.ticker_full)
series_meta.series_name = 'N-month Annualised Change in U.S. CPI Inflation'
series_meta.series_description = 'Calculated series'
return annualised
# Groups:
# food and beverages, housing, apparel, transportation, medical care, recreation, education and communication, and other goods and services
groups_tickers = [
'CPIFABSL', 'CPIHOSSL', 'CPIAPPSL', 'CPITRNSL', 'CPIMEDSL', 'CPIRECSL',
'CPIEDUSL', 'CPIOGSSL'
]
group_df = pandas.DataFrame()
for group in groups_tickers:
level = fetch('F@' + group)
pct_chg = level.pct_change(N_months)
factor = 1. + pct_chg
annualised = pow(factor, 12. / float(N_months))
annualised = 100. * (annualised - 1.)
group_df[group] = annualised
group_df.dropna(axis=0, inplace=True)
if return_type == 'MIN':
out = group_df.min(axis=1)
series_meta.series_name = f'Minimum {N_months}-annualised inflation for major groups in U.S.'
else:
out = group_df.max(axis=1)
series_meta.series_name = f'Maximum {N_months}-annualised inflation for major groups in U.S.'
series_meta.series_description = 'Calculated series'
return out, series_meta
factor = 1. + pct_chg
annualised = pow(factor, 12. / float(N_months))
annualised = 100. * (annualised - 1.)
group_df[group] = annualised
group_df.dropna(axis=0, inplace=True)
if return_type == 'MIN':
out = group_df.min(axis=1)
series_meta.series_name = f'Minimum {N_months}-annualised inflation for major groups in U.S.'
else:
out = group_df.max(axis=1)
series_meta.series_name = f'Maximum {N_months}-annualised inflation for major groups in U.S.'
series_meta.series_description = 'Calculated series'
return out, series_meta
# Push the handler into the UserProvider
user_provider = Providers.UserProvider
user_provider.FunctionMapper['US_INF_DISPERSION'] = inflation_fn
# End of code that should be in a library.
#--------------------------------------------------------------------
# Now we can fetch it.
# Note that since this calculated series uses the US CPI index level, the system
# automatically also creates the CPI series on the database when this is called.
# Call reset_update_time() to force an update of the series while the code is under development.
# Note that this does not affect the series fetched from the external provider.
reset_update_time('U@US_INF_DISPERSION(24, MAX)')
inf = fetch('U@US_INF_DISPERSION(24, MAX)')
quick_plot(inf)
def us_inflation(query_ticker):
if not query_ticker == 'US_CPI_INFLATION':
raise ValueError('Wrong series!')
cpi_index = fetch('F@CPIAUCSL') # All -items, SA
inf = cpi_index.pct_change(12)
return inf
"""
from econ_platform_core import fetch, Providers
from econ_platform.start import quick_plot
#---------------------------------------------------------------------------------------
# Normally, the user-defined series would be handled in a library. Kept here to keep
# Example self-contained.
def us_inflation(query_ticker):
if not query_ticker == 'US_CPI_INFLATION':
raise ValueError('Wrong series!')
cpi_index = fetch('F@CPIAUCSL') # All -items, SA
inf = cpi_index.pct_change(12)
return inf
# Push the handler into the UserProvider
# Assume we are still using the default provider code
user_provider = Providers['U']
user_provider.SeriesMapper['US_CPI_INFLATION'] = us_inflation
# End of code that should be in a library.
#--------------------------------------------------------------------
# Now we can fetch it.
# Note that since this calculated series uses the US CPI index level, the system
# automatically also creates the CPI series on the database when this is called.
inf = fetch('U@US_CPI_INFLATION')
quick_plot(inf)
:param series_meta: econ_platform_core.SeriesMetadata
:return:
"""
if not str(series_meta.ticker_query) in ('GREEK_DRACHMA_OVERNIGHT_AVG',
'GREEK_DRACHMA_OVERNIGHT_EOM'):
raise ValueError('Wrong series!')
full_ticker = tickers.TickerFull(series_meta.ticker_full)
# I will assume that the file is downloaded to the current directory
sheet = pandas.read_excel('drachma_overnight_rate.xls', header=2)
if str(series_meta.ticker_query) == 'GREEK_DRACHMA_OVERNIGHT_AVG':
ser = pandas.Series(sheet['monthly averages'])
ser.index = sheet['Unnamed: 0']
series_meta.series_name = 'Average pre-Euro Greek Drachma Bank Overnight Rate'
series_meta.series_description = 'Monthly average bank overnight rate from Bank of Greece'
# quick_plot(ser)
else:
print(sheet)
raise NotImplementedError('Bam!')
ser.name = str(full_ticker)
return ser
# Push the handler into the UserProvider
user_provider = Providers.UserProvider
user_provider.SeriesMapper['GREEK_DRACHMA_OVERNIGHT_AVG'] = greek_overnight
user_provider.SeriesMapper['GREEK_DRACHMA_OVERNIGHT_EOM'] = greek_overnight
# End of code that should be in a library.
#--------------------------------------------------------------------
# Now we can fetch it.
ser = fetch('U@GREEK_DRACHMA_OVERNIGHT_AVG')
quick_plot(ser)
