Skip to content

Partially revert #5131 #5135

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Nov 5, 2021
Merged

Partially revert #5131 #5135

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
22 changes: 10 additions & 12 deletions pymc/step_methods/metropolis.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -50,59 +50,57 @@


class Proposal:
def __init__(self, s, rng_seed: Optional[int] = None):
def __init__(self, s):
self.s = s
self.rng = np.random.default_rng(rng_seed)


class NormalProposal(Proposal):
def __call__(self, rng: Optional[np.random.Generator] = None):
if rng is None:
rng = self.rng
rng = nr
return rng.normal(scale=self.s)


class UniformProposal(Proposal):
def __call__(self, rng: Optional[np.random.Generator] = None):
if rng is None:
rng = self.rng
rng = nr
return rng.uniform(low=-self.s, high=self.s, size=len(self.s))


class CauchyProposal(Proposal):
def __call__(self, rng: Optional[np.random.Generator] = None):
if rng is None:
rng = self.rng
rng = nr
return rng.standard_cauchy(size=np.size(self.s)) * self.s


class LaplaceProposal(Proposal):
def __call__(self, rng: Optional[np.random.Generator] = None):
if rng is None:
rng = self.rng
rng = nr
size = np.size(self.s)
return (rng.standard_exponential(size=size) - rng.standard_exponential(size=size)) * self.s


class PoissonProposal(Proposal):
def __call__(self, rng: Optional[np.random.Generator] = None):
if rng is None:
rng = self.rng
rng = nr
return rng.poisson(lam=self.s, size=np.size(self.s)) - self.s


class MultivariateNormalProposal(Proposal):
def __init__(self, *args, **kwargs):
super().__init__(*args, **kwargs)
n, m = self.s.shape
def __init__(self, s):
n, m = s.shape
if n != m:
raise ValueError("Covariance matrix is not symmetric.")
self.n = n
self.chol = scipy.linalg.cholesky(self.s, lower=True)
self.chol = scipy.linalg.cholesky(s, lower=True)

def __call__(self, num_draws=None, rng: Optional[np.random.Generator] = None):
if rng is None:
rng = self.rng
rng = nr
if num_draws is not None:
b = rng.normal(size=(self.n, num_draws))
return np.dot(self.chol, b).T
Expand Down