alphacsc.learn_d_z#

alphacsc.learn_d_z(X, n_atoms, n_times_atom, func_d=<function update_d_block>, reg=0.1, lmbd_max='fixed', n_iter=60, random_state=None, n_jobs=1, solver_z='l-bfgs', solver_d_kwargs={}, solver_z_kwargs={}, ds_init=None, sample_weights=None, verbose=10, callback=None, stopping_pobj=None)#

Univariate Convolutional Sparse Coding.

Parameters:
Xarray, shape (n_trials, n_times)

The data on which to perform CSC.

n_atomsint

The number of atoms to learn.

n_times_atomint

The support of the atom.

func_dcallable

The function to update the atoms.

regfloat

The regularization parameter

lmbd_max‘fixed’ | ‘scaled’ | ‘per_atom’ | ‘shared’
If not fixed, adapt the regularization rate as a ratio of lambda_max:
  • ‘scaled’: the regularization parameter is fixed as a ratio of its maximal value at init i.e. reg_used = reg * lmbd_max(uv_init)

  • ‘shared’: the regularization parameter is set at each iteration as a ratio of its maximal value for the current dictionary estimate i.e. reg_used = reg * lmbd_max(uv_hat)

  • ‘per_atom’: the regularization parameter is set per atom and at each iteration as a ratio of its maximal value for this atom i.e. reg_used[k] = reg * lmbd_max(uv_hat[k])

n_iterint

The number of coordinate-descent iterations.

random_stateint | None

The random state.

n_jobsint

The number of parallel jobs.

solver_zstr

The solver to use for the z update. Options are ‘l-bfgs’ (default) | ‘ista’ | ‘fista’

solver_d_kwargsdict

Additional keyword arguments to provide to update_d

solver_z_kwargsdict

Additional keyword arguments to pass to update_z

ds_initstr or array, shape (n_atoms, n_times_atom)

The initial atoms or an initialization scheme in {‘chunk’ | ‘random’}.

sample_weightsarray, shape (n_trials, n_times)

The weights in the alphaCSC problem. Should be None when using vanilla CSC.

verboseint

The verbosity level.

callbackfunc

A callback function called at the end of each loop of the coordinate descent.

Returns:
pobjlist

The objective function value at each step of the coordinate descent.

timeslist

The cumulative time for each iteration of the coordinate descent.

d_hatarray, shape (n_atoms, n_times)

The estimated atoms.

z_hatarray, shape (n_atoms, n_trials, n_times - n_times_atom + 1)

The sparse activation matrix.

regfloat

Regularization parameter used.