Home       Trees       Indices       Help   
Modular toolkit for Data Processing MDP
Package mdp :: Package nodes :: Class FastICANode
[hide private]
[frames] | no frames]

Class FastICANode

Perform Independent Component Analysis using the FastICA algorithm. Note that FastICA is a batch-algorithm. This means that it needs all input data before it can start and compute the ICs. The algorithm is here given as a Node for convenience, but it actually accumulates all inputs it receives. Remember that to avoid running out of memory when you have many components and many time samples.

FastICA does not support the telescope mode (the convergence criterium is not robust in telescope mode).

Reference: Aapo Hyvarinen (1999). Fast and Robust Fixed-Point Algorithms for Independent Component Analysis IEEE Transactions on Neural Networks, 10(3):626-634.

Internal variables of interest

self.white
The whitening node used for preprocessing.
self.filters
The ICA filters matrix (this is the transposed of the projection matrix after whitening).
self.convergence
The value of the convergence threshold.

History:

Instance Methods [hide private]
 
__init__(self, approach='defl', g='pow3', guess=None, fine_g='pow3', mu=1, sample_size=1, fine_tanh=1, fine_gaus=1, max_it=5000, max_it_fine=100, failures=5, coarse_limit=None, limit=0.001, verbose=False, whitened=False, white_comp=None, white_parm=None, input_dim=None, dtype=None)
Input arguments:
 
_get_rsamples(self, X)
 
core(self, data)
This is the core routine of the ICANode. Each subclass must define this function to return the achieved convergence value. This function is also responsible for setting the ICA filters matrix self.filters. Note that the matrix self.filters is applied to the right of the matrix containing input data. This is the transposed of the matrix defining the linear transformation.

Inherited from unreachable.ProjectMatrixMixin: get_projmatrix, get_recmatrix

Inherited from unreachable.newobject: __long__, __native__, __nonzero__, __unicode__, next

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

    Inherited from ICANode
 
_execute(self, x)
 
_inverse(self, y)
 
_set_input_dim(self, n)
 
_stop_training(self)
Whiten data if needed and call the 'core' routine to perform ICA. Take care of telescope-mode if needed.
 
execute(self, x)
Process the data contained in x.
 
inverse(self, y)
Invert y.
 
stop_training(self)
Whiten data if needed and call the 'core' routine to perform ICA. Take care of telescope-mode if needed.
    Inherited from Cumulator
 
_train(self, *args)
Collect all input data in a list.
 
train(self, *args)
Collect all input data in a list.
    Inherited from Node
 
__add__(self, other)
 
__call__(self, x, *args, **kwargs)
Calling an instance of Node is equivalent to calling its execute method.
 
__repr__(self)
repr(x)
 
__str__(self)
str(x)
 
_check_input(self, x)
 
_check_output(self, y)
 
_check_train_args(self, x, *args, **kwargs)
 
_get_supported_dtypes(self)
Return the list of dtypes supported by this node.
 
_get_train_seq(self)
 
_if_training_stop_training(self)
 
_pre_execution_checks(self, x)
This method contains all pre-execution checks.
 
_pre_inversion_checks(self, y)
This method contains all pre-inversion checks.
 
_refcast(self, x)
Helper function to cast arrays to the internal dtype.
 
_set_dtype(self, t)
 
_set_output_dim(self, n)
 
copy(self, protocol=None)
Return a deep copy of the node.
 
get_current_train_phase(self)
Return the index of the current training phase.
 
get_dtype(self)
Return dtype.
 
get_input_dim(self)
Return input dimensions.
 
get_output_dim(self)
Return output dimensions.
 
get_remaining_train_phase(self)
Return the number of training phases still to accomplish.
 
get_supported_dtypes(self)
Return dtypes supported by the node as a list of dtype objects.
 
has_multiple_training_phases(self)
Return True if the node has multiple training phases.
 
is_training(self)
Return True if the node is in the training phase, False otherwise.
 
save(self, filename, protocol=-1)
Save a pickled serialization of the node to filename. If filename is None, return a string.
 
set_dtype(self, t)
Set internal structures' dtype.
 
set_input_dim(self, n)
Set input dimensions.
 
set_output_dim(self, n)
Set output dimensions.
Static Methods [hide private]
    Inherited from Node
 
is_invertible()
Return True if the node can be inverted, False otherwise.
 
is_trainable()
Return True if the node can be trained, False otherwise.
Properties [hide private]

Inherited from object: __class__

    Inherited from Node
  _train_seq
List of tuples:
  dtype
dtype
  input_dim
Input dimensions
  output_dim
Output dimensions
  supported_dtypes
Supported dtypes
Method Details [hide private]

__init__(self, approach='defl', g='pow3', guess=None, fine_g='pow3', mu=1, sample_size=1, fine_tanh=1, fine_gaus=1, max_it=5000, max_it_fine=100, failures=5, coarse_limit=None, limit=0.001, verbose=False, whitened=False, white_comp=None, white_parm=None, input_dim=None, dtype=None)
(Constructor)

 

Input arguments:

General:

whitened -- Set whitened == True if input data are already whitened.
Otherwise the node will whiten the data itself
white_comp -- If whitened == False, you can set 'white_comp' to the
number of whitened components to keep during the calculation (i.e., the input dimensions are reduced to white_comp by keeping the components of largest variance).
white_parm -- a dictionary with additional parameters for whitening.
It is passed directly to the WhiteningNode constructor. Ex: white_parm = { 'svd' : True }

limit -- convergence threshold.

Specific for FastICA:

approach -- Approach to use. Possible values are:
'defl' --> deflation 'symm' --> symmetric
g -- Nonlinearity to use. Possible values are:
'pow3' --> x^3 'tanh' --> tanh(fine_tanh*x) 'gaus' --> x*exp(-fine_gaus*x^2/2) 'skew' --> x^2 (for skewed signals)
fine_g -- Nonlinearity for fine tuning. Possible values
are the same as for 'g'. Set it to None to disable fine tuning.
mu -- Step size. If mu != 1, a stabilization procedure is used:
the value of mu can momentarily be halved if the algorithm is stuck between two points (this is called a stroke). Also if there is no convergence before half of the maximum number of iterations has been reached then mu will be halved for the rest of the rounds.
sample_size -- Percentage of samples used in one iteration. If
sample_size < 1, samples are chosen in random order.
coarse_limit -- initial convergence threshold, to switch to
fine_g function (i.e. linear to non-linear) even before reaching the limit and final tuning. Set it to a value higher than limit to be in effect.

fine_tanh -- parameter for 'tanh' nonlinearity fine_gaus -- parameter for 'gaus' nonlinearity

guess -- initial guess for the mixing matrix (ignored if None)

max_it -- maximum number of iterations

max_it_fine -- maximum number of iterations for fine tuning

failures -- maximum number of failures to allow in deflation mode
Overrides: object.__init__

_get_rsamples(self, X)

 

core(self, data)

 
This is the core routine of the ICANode. Each subclass must define this function to return the achieved convergence value. This function is also responsible for setting the ICA filters matrix self.filters. Note that the matrix self.filters is applied to the right of the matrix containing input data. This is the transposed of the matrix defining the linear transformation.
Overrides: ICANode.core
(inherited documentation)

   Home       Trees       Indices       Help   
Modular toolkit for Data Processing MDP
Generated by Epydoc 3.0.1 on Tue Mar 8 12:39:48 2016 http://epydoc.sourceforge.net