mofaflex.priors.Constant

class mofaflex.priors.Constant(const_values)

Pseudo-prior that always returns the same constant values.

When used for weights, this can be used to project new data into an already existing latent space.

Parameters:

const_values (Mapping[str, DataFrame]) – The constant values to use for each group/view. Must have factors in columns and samples/features in rows. The number of factors must match the configuration of the model this prior is being used with. Sample/feature names must be a superset of the sample/feature names of the data this prior is being used with.