PnPULA#
- class cuqi.experimental.mcmc.PnPULA(target=None, scale=1.0, smoothing_strength=0.1, **kwargs)#
Plug-and-Play Unadjusted Langevin algorithm (PnP-ULA) (Laumont et al., 2022)
Samples a smoothed target distribution given its smoothed logpdf gradient based on Langevin diffusion dL_t = dW_t + 1/2*Nabla target.logd(L_t)dt, where W_t is a dim-dimensional standard Brownian motion. It targets a differentiable density (partially) smoothed by a convolution with Gaussian kernel with zero mean and smoothing_strength variance. The smoothed target density can be made arbitrarily closed to the true unsmoothed target density.
For more details see: Laumont, R., Bortoli, V. D., Almansa, A., Delon, J., Durmus, A., & Pereyra, M. (2022). Bayesian imaging using plug & play priors: when Langevin meets Tweedie. SIAM Journal on Imaging Sciences, 15(2), 701-737.
- Parameters:
target (cuqi.distribution.Distribution) – The target distribution to sample. The target distribution result from a differentiable likelihood and prior of type RestorationPrior.
initial_point (ndarray) – Initial parameters. Optional
scale (float) – The Langevin diffusion discretization time step (In practice, a scale of 1/L, where L is the Lipschitz of the gradient of the log target density is recommended but not guaranteed to be the optimal choice).
smoothing_strength (float) – This parameter controls the smoothing strength of PnP-ULA.
- callbackcallable, Optional
If set this function will be called after every sample. The signature of the callback function is callback(sample, sample_index), where sample is the current sample and sample_index is the index of the sample. An example is shown in demos/demo31_callback.py.
# TODO: update demo once sampler merged
- __init__(target=None, scale=1.0, smoothing_strength=0.1, **kwargs)#
Initializer for abstract base class for all samplers.
Any subclassing samplers should simply store input parameters as part of the __init__ method.
The actual initialization of the sampler should be done in the _initialize method.
- Parameters:
target (cuqi.density.Density) – The target density.
initial_point (array-like, optional) – The initial point for the sampler. If not given, the sampler will choose an initial point.
callback (callable, optional) – A function that will be called after each sample is drawn. The function should take two arguments: the sample and the index of the sample. The sample is a 1D numpy array and the index is an integer. The callback function is useful for monitoring the sampler during sampling.
Methods
__init__
([target, scale, smoothing_strength])Initializer for abstract base class for all samplers.
Return the history of the sampler.
Return the samples.
Return the state of the sampler.
Initialize the sampler by setting and allocating the state and history before sampling starts.
load_checkpoint
(path)Load the state of the sampler from a file.
Re-initialize the sampler.
sample
(Ns[, batch_size, sample_path])Sample Ns samples from the target density.
save_checkpoint
(path)Save the state of the sampler to a file.
set_history
(history)Set the history of the sampler.
set_state
(state)Set the state of the sampler.
step
()Perform one step of the sampler by transitioning the current point to a new point according to the sampler's transition kernel.
tune
(skip_len, update_count)Tune the parameters of the sampler.
Validate the target is compatible with the sampler.
warmup
(Nb[, tune_freq])Warmup the sampler by drawing Nb samples.
Attributes