gmisclib.mcmc_newlogger.logger

Class logger

Instance Methods

__init__(self, fd, huge=1e+30)
Create a data log.

source code

add(self, uid, prms, map, logp, iter, extra=None, reason=None)
Adds a set of parameters to the log file.

source code

close(self) (Inherited from gmisclib.mcmc_newlogger.logger_base)

source code

comment(self, comment) (Inherited from gmisclib.mcmc_newlogger.logger_base)

source code

flush(self) (Inherited from gmisclib.mcmc_newlogger.logger_base)

source code

header(self, k, v) (Inherited from gmisclib.mcmc_newlogger.logger_base)

source code

headers(self, hdict) (Inherited from gmisclib.mcmc_newlogger.logger_base)

source code

Inherited from object: __delattr__, __format__, __getattribute__, __hash__, __new__, __reduce__, __reduce_ex__, __repr__, __setattr__, __sizeof__, __str__, __subclasshook__

Properties

Inherited from object: __class__

Method Details

init(self, fd, huge=1e+30)
(Constructor)

source code

Create a data log.

Parameters:

fd (file) - Where to log the results
huge (normally float, but could be numpy.ndarray) - sets a threshold for throwing WildNumber. If the absolute value of any parameter gets bigger than huge, something is assumed to be wrong.

Overrides: object.__init__

add(self, uid, prms, map, logp, iter, extra=None, reason=None)

source code

Adds a set of parameters to the log file.

Parameters:

uid (string) - which subset of parameters,
prms (numpy.ndarray) - a vector of the actual numbers,
map (dict) - a dictionary mapping from parameter names to an index into the prms array.
logp (float) - how good is the fit? log(probability)
iter (int) - which iteration is it?
extra (None or a dictionary.) - any extra information desired.

Class logger

__init__(self, fd, huge=1e+30) (Constructor)

add(self, uid, prms, map, logp, iter, extra=None, reason=None)

init(self, fd, huge=1e+30)
(Constructor)