Class that amends ChainNode with a mapper-like interface.
ChainMapper supports sequential training of a mapper chain, as well as reverse-mapping and mapping of single samples.
Notes
Available conditional attributes:
(Conditional attributes enabled by default suffixed with +)
Methods
append(node) | Append a node to the chain. | ||
forward(ds) | |||
forward1(data) | Forward data or datasets through the chain. | ||
generate(ds[, startnode]) |
|
||
get_postproc() | Returns the post-processing node or None. | ||
get_space() | Query the processing space name of this node. | ||
reset() | |||
reverse(data) | Reverse-maps data or datasets through the chain (backwards). | ||
reverse1(data) | Reverse-maps data or datasets through the chain (backwards). | ||
set_postproc(node) | Assigns a post-processing node Set to None to disable postprocessing. | ||
set_space(name) | Set the processing space name of this node. | ||
train(dataset) | Train the mapper chain sequentially. | ||
untrain() | Untrain all embedded mappers. |
Parameters: | nodes: list :
enable_ca : None or list of str
disable_ca : None or list of str
space: str, optional :
postproc : Node instance, optional
descr : str
|
---|
Methods
append(node) | Append a node to the chain. | ||
forward(ds) | |||
forward1(data) | Forward data or datasets through the chain. | ||
generate(ds[, startnode]) |
|
||
get_postproc() | Returns the post-processing node or None. | ||
get_space() | Query the processing space name of this node. | ||
reset() | |||
reverse(data) | Reverse-maps data or datasets through the chain (backwards). | ||
reverse1(data) | Reverse-maps data or datasets through the chain (backwards). | ||
set_postproc(node) | Assigns a post-processing node Set to None to disable postprocessing. | ||
set_space(name) | Set the processing space name of this node. | ||
train(dataset) | Train the mapper chain sequentially. | ||
untrain() | Untrain all embedded mappers. |
Reverse-maps data or datasets through the chain (backwards).
See Mapper for more information.
Reverse-maps data or datasets through the chain (backwards).
See Mapper for more information.
Train the mapper chain sequentially.
The training dataset is used to train the first mapper. Afterwards it is forward-mapped by this (now trained) mapper and the transformed dataset and then used to train the next mapper. This procedure is done till all mappers are trained.
Parameters: | dataset: `Dataset` : |
---|
Untrain all embedded mappers.