qif.channel¶
Channels.
Channel composition. |
- qif.channel.assert_proper(*args, **kwargs)¶
Overloaded function.
assert_proper(C: Mat<double>) -> None
assert_proper(C: Mat<rat>) -> None
- qif.channel.deterministic(*args, **kwargs)¶
Overloaded function.
deterministic(map: Callable[[int], int], n_rows: int, n_cols: int, type: class[double] = 1) -> Mat<double>
deterministic(map: Callable[[int], int], n_rows: int, n_cols: int, type: class[fraction] = 1) -> Mat<rat>
- qif.channel.factorize(*args, **kwargs)¶
Overloaded function.
factorize(A: Mat<double>, B: Mat<double>, col_stoch: bool = False) -> Mat<double>
factorize(A: Mat<rat>, B: Mat<rat>, col_stoch: bool = False) -> Mat<rat>
- qif.channel.hyper(*args, **kwargs)¶
Overloaded function.
hyper(C: Mat<double>, pi: Row<double>) -> Tuple[Row<double>, Mat<double>]
hyper(C: Mat<rat>, pi: Row<rat>) -> Tuple[Row<rat>, Mat<rat>]
- qif.channel.identity(*args, **kwargs)¶
Overloaded function.
identity(n_rows: int, type: class[double] = 1) -> Mat<double>
identity(n_rows: int, type: class[fraction] = 1) -> Mat<rat>
- qif.channel.is_proper(*args, **kwargs)¶
Overloaded function.
is_proper(C: Mat<double>, mrd: float = 2.220446049250313e-14) -> bool
is_proper(C: Mat<rat>, mrd: mppp::rational<1> = Fraction(0, 1)) -> bool
- qif.channel.iterative_bayesian_update(*args, **kwargs)¶
Overloaded function.
iterative_bayesian_update(C: Mat<double>, out: Row<double>, start: Row<double> = array([], dtype=float64), max_diff: float = 1e-06, max_iter: int = 0) -> Tuple[Row<double>, int]
iterative_bayesian_update(C: Mat<rat>, out: Row<rat>, start: Row<rat>, max_diff: mppp::rational<1> = 1e-06, max_iter: int = 0) -> Tuple[Row<rat>, int]
- qif.channel.left_factorize(*args, **kwargs)¶
Overloaded function.
left_factorize(A: Mat<double>, B: Mat<double>, col_stoch: bool = False) -> Mat<double>
left_factorize(A: Mat<rat>, B: Mat<rat>, col_stoch: bool = False) -> Mat<rat>
- qif.channel.no_interference(*args, **kwargs)¶
Overloaded function.
no_interference(n_rows: int, n_cols: int = 1, type: class[double] = 1) -> Mat<double>
no_interference(n_rows: int, n_cols: int = 1, type: class[fraction] = 1) -> Mat<rat>
- qif.channel.normalize(*args, **kwargs)¶
Overloaded function.
normalize(C: Mat<double>) -> Mat<double>
normalize(C: Mat<rat>) -> Mat<rat>
- qif.channel.posterior(*args, **kwargs)¶
Overloaded function.
posterior(C: Mat<double>, pi: Row<double>, col: int) -> Row<double>
posterior(C: Mat<rat>, pi: Row<rat>, col: int) -> Row<rat>
- qif.channel.posteriors(*args, **kwargs)¶
Overloaded function.
posteriors(C: Mat<double>, pi: Row<double> = array([], dtype=float64)) -> Mat<double>
posteriors(C: Mat<rat>, pi: Row<rat>) -> Mat<rat>
- qif.channel.randu(*args, **kwargs)¶
Overloaded function.
randu(n_rows: int, n_cols: int = 0, type: class[double] = 1) -> Mat<double>
randu(n_rows: int, n_cols: int = 0, type: class[fraction] = 1) -> Mat<rat>
- qif.channel.reduced(*args, **kwargs)¶
Overloaded function.
reduced(C: Mat<double>) -> Mat<double>
reduced(C: Mat<rat>) -> Mat<rat>
- qif.channel.sample(*args, **kwargs)¶
Overloaded function.
sample(C: Mat<double>, pi: Row<double>) -> Tuple[int, int]
sample(C: Mat<rat>, pi: Row<rat>) -> Tuple[int, int]
sample(C: Mat<double>, pi: Row<double>, n_samples: int) -> Mat<uint>
sample(C: Mat<rat>, pi: Row<rat>, n_samples: int) -> Mat<uint>
- qif.channel.sum_column_min(*args, **kwargs)¶
Overloaded function.
sum_column_min(C: Mat<double>) -> float
sum_column_min(C: Mat<rat>) -> mppp::rational<1>