gnes.score_fn.base module

class gnes.score_fn.base.BaseScoreFn(context=None, *args, **kwargs)

Bases: gnes.base.TrainableBase

Base score function. A score function must implement __call__ method

new_score(*, operands=(), **kwargs)

train(*args, **kwargs)

Train the model, need to be overrided

warn_unnamed = False

class gnes.score_fn.base.CombinedScoreFn(score_mode='multiply', *args, **kwargs)

Bases: gnes.score_fn.base.BaseScoreFn

Combine multiple scores into one score, defaults to ‘multiply’

Parameters:score_mode (str) – specifies how the computed scores are combined

post_init()

Declare class attributes/members that can not be serialized in standard way

supported_ops

train(*args, **kwargs)

Train the model, need to be overrided

class gnes.score_fn.base.ModifierScoreFn(modifier='none', factor=1.0, factor_name='GivenConstant', *args, **kwargs)

Bases: gnes.score_fn.base.BaseScoreFn

Modifier to apply to the value score = modifier(factor * value)

post_init()

Declare class attributes/members that can not be serialized in standard way

supported_ops

train(*args, **kwargs)

Train the model, need to be overrided

class gnes.score_fn.base.ScoreOps

Bases: object

abs = <gnes.score_fn.base.ModifierScoreFn object>

avg = <gnes.score_fn.base.CombinedScoreFn object>

ln = <gnes.score_fn.base.ModifierScoreFn object>

ln1p = <gnes.score_fn.base.ModifierScoreFn object>

ln2p = <gnes.score_fn.base.ModifierScoreFn object>

log = <gnes.score_fn.base.ModifierScoreFn object>

log1p = <gnes.score_fn.base.ModifierScoreFn object>

log2p = <gnes.score_fn.base.ModifierScoreFn object>

max = <gnes.score_fn.base.CombinedScoreFn object>

min = <gnes.score_fn.base.CombinedScoreFn object>

multiply = <gnes.score_fn.base.CombinedScoreFn object>

none = <gnes.score_fn.base.ModifierScoreFn object>

reciprocal = <gnes.score_fn.base.ModifierScoreFn object>

reciprocal1p = <gnes.score_fn.base.ModifierScoreFn object>

sqrt = <gnes.score_fn.base.ModifierScoreFn object>

square = <gnes.score_fn.base.ModifierScoreFn object>

sum = <gnes.score_fn.base.CombinedScoreFn object>

gnes.score_fn.base.get_unary_score(value, **kwargs)