Skip to content

Commit

Permalink
release 0.1.4
Browse files Browse the repository at this point in the history
  • Loading branch information
@gpitt71
gpitt71 committed Aug 24, 2022
1 parent 5738d7b commit 5eaee5c
Show file tree
Hide file tree
Showing 4 changed files with 517 additions and 509 deletions.
2 changes: 1 addition & 1 deletion __init__.py
View file
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
__version__ = "0.1.3"
__version__ = "0.1.4"
__doc__ = """GEMAct is an **actuarial package**,
based on the collective risk theory framework,
that offers actuarial scientists and practitioners a comprehensive set of tools for
Expand Down
72 changes: 36 additions & 36 deletions copulas.py
View file
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@
import time
import numpy as np
from scipy.stats import multivariate_normal, norm, t, gamma, uniform, multivariate_t, logser
from . import helperfunctions as hf
import time # used for setting random number generator seed if None
from twiggy import quick_setup, log

quick_setup()
Expand Down Expand Up @@ -57,13 +57,13 @@ def cdf(self, x):
return 0
else:
return (np.sum(np.minimum(x, 1) ** (-self.par) - 1) + 1) ** -(1 / self.par)

capital_n = len(x)
output = np.array([0.] * capital_n)
index = ~np.array(x <= 0).any(axis=1)
output[index] = (np.sum(np.minimum(x[index, :], 1) ** (-self.par) - 1, axis=1) + 1) ** (-1 / self.par)
return output

def rvs(self, size=1, random_state=None):
"""
Random variates.
Expand All @@ -74,7 +74,7 @@ def rvs(self, size=1, random_state=None):
:type random_state: ``int``
:return: Random variates.
:rtype: ``numpy.float64`` or ``numpy.ndarray``
:rtype: ``numpy.float64`` or ``numpy.ndarray``
"""
random_state = int(time.time()) if random_state is None else random_state
assert isinstance(random_state, int), logger.error("%r is not an integer" % random_state)
Expand All @@ -85,9 +85,9 @@ def rvs(self, size=1, random_state=None):
logger.error('Please make sure random_state is provided correctly')
raise

gamma_sim = gamma.rvs(1/self.par, size=[size, 1], random_state=random_state)
exp_sim = gamma.rvs(1, size=[size, self.dim], random_state=random_state+2)
output = (1 + exp_sim/gamma_sim)**(-1/self.par)
gamma_sim = gamma.rvs(1 / self.par, size=[size, 1], random_state=random_state)
exp_sim = gamma.rvs(1, size=[size, self.dim], random_state=random_state + 2)
output = (1 + exp_sim / gamma_sim) ** (-1 / self.par)
return output


Expand Down Expand Up @@ -151,7 +151,7 @@ def cdf(self, x):
1 + np.prod(np.exp(-self.par * np.minimum(x[index, :], 1)) - 1, axis=1) / (
np.exp(-self.par) - 1) ** (d - 1))
return output

def rvs(self, size=1, random_state=None):
"""
Random variates.
Expand All @@ -162,7 +162,7 @@ def rvs(self, size=1, random_state=None):
:type random_state: ``int``
:return: Random variates.
:rtype: ``numpy.float64`` or ``numpy.ndarray``
:rtype: ``numpy.float64`` or ``numpy.ndarray``
"""
random_state = int(time.time()) if random_state is None else random_state
assert isinstance(random_state, int), logger.error("%r is not an integer" % random_state)
Expand All @@ -173,9 +173,9 @@ def rvs(self, size=1, random_state=None):
logger.error('Please make sure random_state is provided correctly')
raise

logarithmic_sim = logser.rvs(1-np.exp(-self.par), size=[size, 1], random_state=random_state)
logarithmic_sim = logser.rvs(1 - np.exp(-self.par), size=[size, 1], random_state=random_state)
exp_sim = gamma.rvs(1, size=[size, self.dim], random_state=random_state)
output = -1/self.par * np.log(1 + np.exp(-exp_sim/logarithmic_sim)*(np.exp(-self.par) - 1))
output = -1 / self.par * np.log(1 + np.exp(-exp_sim / logarithmic_sim) * (np.exp(-self.par) - 1))
return output


Expand Down Expand Up @@ -245,7 +245,7 @@ def rvs(self, size=1, random_state=None):
:type random_state: ``int``
:return: Random variates.
:rtype: ``numpy.float64`` or ``numpy.ndarray``
:rtype: ``numpy.float64`` or ``numpy.ndarray``
"""
random_state = int(time.time()) if random_state is None else random_state
assert isinstance(random_state, int), logger.error("%r is not an integer" % random_state)
Expand All @@ -256,13 +256,14 @@ def rvs(self, size=1, random_state=None):
logger.error('Please make sure random_state is provided correctly.')
raise

a_ = 1/self.par
uniform_sim = (uniform.rvs(size=[size, 1], random_state=random_state)-0.5)*np.pi
a_ = 1 / self.par
uniform_sim = (uniform.rvs(size=[size, 1], random_state=random_state) - 0.5) * np.pi
exp_sim = gamma.rvs(1, size=[size, 1], random_state=random_state)
stable_sim = np.sin(a_*(np.pi/2 + uniform_sim)) / \
np.cos(uniform_sim)**(1/a_) * (np.cos(uniform_sim - a_*(np.pi/2 + uniform_sim))/exp_sim)**((1 - a_)/a_)
exp_sim = gamma.rvs(1, size=[size, self.dim], random_state=random_state+2)
output = np.exp(-(exp_sim/stable_sim)**(1/self.par))
stable_sim = np.sin(a_ * (np.pi / 2 + uniform_sim)) / np.cos(
uniform_sim) ** (1 / a_) * (
np.cos(uniform_sim - a_ * (np.pi / 2 + uniform_sim)) / exp_sim) ** ((1 - a_) / a_)
exp_sim = gamma.rvs(1, size=[size, self.dim], random_state=random_state + 2)
output = np.exp(-(exp_sim / stable_sim) ** (1 / self.par))
return output


Expand Down Expand Up @@ -305,7 +306,7 @@ def cdf(self, x):
:rtype: ``numpy.ndarray``
"""
return multivariate_normal.cdf(norm.ppf(x), cov=self.corr)

def rvs(self, size=1, random_state=None):
"""
Random variates.
Expand All @@ -316,7 +317,7 @@ def rvs(self, size=1, random_state=None):
:type random_state: ``int``
:return: Random variates.
:rtype: ``numpy.float64`` or ``numpy.ndarray``
:rtype: ``numpy.float64`` or ``numpy.ndarray``
"""
random_state = int(time.time()) if random_state is None else random_state
assert isinstance(random_state, int), logger.error("%r has to be an integer" % random_state)
Expand All @@ -332,15 +333,15 @@ def rvs(self, size=1, random_state=None):
cov=self.corr,
size=size,
random_state=random_state
)
)
return norm.cdf(sim)


# TCopula
class TCopula:
"""
T-Student copula.
:param corr: Correlation matrix.
:type corr: ``numpy.ndarray``
:param df: Degree of freedom.
Expand Down Expand Up @@ -369,7 +370,7 @@ def corr(self, value):
if not np.allclose(np.diagonal(value), np.ones(value.shape[0])):
raise ValueError('%r is not a correlation matrix' % value)
self.__corr = value

@property
def df(self):
return self.__df
Expand All @@ -390,21 +391,21 @@ def cdf(self, x, tolerance=1e-4, max_evaluations=1e+6, n_repetitions=30):
:param x: Array with shape (N, d) where N is the number of points and d the dimension.
:type x: ``numpy.ndarray``
:param tolerance: Tolerance threshold in cdf computation.
:type tolerance: ``float``
:param max_evaluations: Maximum number of iterations.
:type max_evaluations: ``int``
:param n_repetitions: Array with shape (N, d) where N is the number of points and d the dimension.
:type n_repetitions: ``int``
:return: Cumulative distribution function in x.
:rtype: ``numpy.ndarray``
"""
try:
x = x.reshape(-1, self.dim)
except Exception:
logger.error('x has wrong dimension')

q = t(self.df).ppf(np.ravel(x)).reshape(x.shape)
prob, err = np.empty(x.shape[0]), np.empty(x.shape[0])
for j in range(x.shape[0]):
prob[j], err[j] = hf.multivariate_t_cdf(q[j, ], self.corr, self.df, tolerance, max_evaluations, n_repetitions)
q = t(self.df).ppf(x)
(prob, err) = hf.multivariate_t_cdf(q, self.corr, self.df, tolerance, max_evaluations, n_repetitions)
self.__error_cdf = err
return prob

def rvs(self, size=1, random_state=None):
"""
Random variates.
Expand All @@ -415,7 +416,7 @@ def rvs(self, size=1, random_state=None):
:type random_state: ``int``
:return: Random variates.
:rtype: ``numpy.float64`` or ``numpy.ndarray``
:rtype: ``numpy.float64`` or ``numpy.ndarray``
"""
random_state = int(time.time()) if random_state is None else random_state
assert isinstance(random_state, int), logger.error("%r has to be an integer" % random_state)
Expand All @@ -431,6 +432,5 @@ def rvs(self, size=1, random_state=None):
shape=self.corr,
size=size,
random_state=random_state
)
)
return t.cdf(sim, df=self.df)

Loading

0 comments on commit 5eaee5c

Please sign in to comment.