Financial Modelling in Python

P1: JYS

c06 JWBK378-Fletcher May 12, 2009 18:55 Printer: Yet to come

82 Financial Modelling in Python

Dec-29,

1000000.000000, JPY, [2008-Dec-29, 2009-Jun-29], basis

30360, 2009-

Jun-29,

1000000.000000, JPY, [2009-Jun-29, 2009-Dec-29], basis

30360, 2009-

Dec-29,

1000000.000000, JPY, [2009-Dec-29, 2010-Jun-29], basis

30360, 2010-

Jun-29,

1000000.000000, JPY, [2010-Jun-29, 2010-Dec-29], basis

30360, 2010-

Dec-29,

1000000.000000, JPY, [2010-Dec-29, 2011-Jun-29], basis

30360, 2011-

Jun-29,

1000000.000000, JPY, [2011-Jun-29, 2011-Dec-29], basis

30360, 2011-

Dec-29,

1000000.000000, JPY, [2011-Dec-29, 2012-Jun-29], basis

30360, 2012-

Jun-29,

1000000.000000, JPY, [2012-Jun-29, 2012-Dec-31], basis

30360, 2012-

Dec-31,

1000000.000000, JPY, [2012-Dec-31, 2013-Jun-28], basis

30360, 2013-

Jun-28,

1000000.000000, JPY, [2013-Jun-28, 2013-Dec-30], basis

30360, 2013-

Dec-30,

1000000.000000, JPY, [2013-Dec-30, 2014-Jun-30], basis

30360, 2014-

Jun-30,

1000000.000000, JPY, [2014-Jun-30, 2014-Dec-29], basis

30360, 2014-

Dec-29,

1000000.000000, JPY, [2014-Dec-29, 2015-Jun-29], basis

30360, 2015-

Jun-29,

1000000.000000, JPY, [2015-Jun-29, 2015-Dec-29], basis

30360, 2015-

Dec-29,

1000000.000000, JPY, [2015-Dec-29, 2016-Jun-29], basis

30360, 2016-

Jun-29,

1000000.000000, JPY, [2016-Jun-29, 2016-Dec-29], basis

30360, 2016-

Dec-29,

1000000.000000, JPY, [2016-Dec-29, 2017-Jun-29], basis

30360, 2017-

Jun-29,

6.3 ADJUVANTS

A financial payoff can depend on constants that vary over time, such as a gearing or barrier

level. The class adjuvant

table in module ppf.core.adjuvant table provides a

simple structure for storing these time-dependent constants. The class is constructed by taking

a list of keys, or equivalently variable names, the dates on which they apply, and the values.

Access to a particular constant at a specific time is provided via the function call operator.

class adjuvant table:

def

init (self, keys, dates, values):

if len(values.shape) <>2:

raise RuntimeError, "expected 2d array of values"

if len(keys) <> values.shape[0] or len(dates) <>

values.shape[1]:

raise RuntimeError, "incorrect size of values array"

P1: JYS

c06 JWBK378-Fletcher May 12, 2009 18:55 Printer: Yet to come

Data Model 83

self. table = {}

i=0

for key in keys:

elem = {}

j=0

for dt in dates:

elem[dt.julian

day()] = values[i][j]

j+=1

self.

table[key] = elem

i+=1

def

call (self, key, dt):

if self.

table.has key(key):

elem = self.

table.get(key)

if elem.has

key(dt.julian day()):

return elem.get(dt.julian

day())

else:

raise RuntimeError, \

"unable to find date in adjuvant table"+" dt = "+str(dt)

else:

raise RuntimeError, "unable to find key in adjuvant table"

def

str (self):

return self.

table. str ()

A high-level generator for adjuvant tables is provided in the ppf.core.generate

adjuvant table module.

def generate adjuvant table(

keys

, tenors

, values

, start

date

, roll

period = 6

, roll

duration = ppf.date time.months

, holiday

centres = None

, shift

method = ppf.date time.shift convention.modified following

, *arguments

, **keywords):

if len(values.shape) <>2:

raise RuntimeError, "expected 2d array of values"

if len(keys) <> values.shape[0] or len(tenors) <> values.shape[1]:

raise RuntimeError, "incorrect size of values array"

from ppf.date

time import days

shift = ppf.date

time.shift

day=0

dates = []

indices = []

cnt=0

start = start

date