MOFAOptions {SMRUCC.genomics.Analysis.Microarray.MultiOmics.MOFA}

.NET clr documentation

MOFAOptions

Description

Declare

            
# namespace SMRUCC.genomics.Analysis.Microarray.MultiOmics.MOFA
export class MOFAOptions {
   # ELBO convergence tolerance (relative change)
   ConvergenceTolerance: double;
   # Minimum fraction of variance explained for a factor to be retained. Factors below this threshold are pruned. (Paper uses 2%.)
   DropFactorThreshold: double;
   # Number of initial "burn-in" iterations without sparsity (paper Sec. 5)
   DropIterations: integer;
   # Maximum number of variational iterations
   MaxIterations: integer;
   # Initial number of factors K (will be pruned during training)
   NumFactors: integer;
   # Print ELBO every N iterations when verbose
   PrintEvery: integer;
   # Random seed for reproducibility
   Seed: Nullable`1;
   # Prior precision for the spike distribution (small value → strong sparsity). Set to 0 to disable spike-and-slab feature sparsity (use ARD only).
   SpikePrecision: double;
   # Whether to standardize each view before training (recommended)
   StandardizeViews: boolean;
   # Verbose output during training
   Verbose: boolean;
}

        

.NET clr type reference tree

1. use by property member Seed: Nullable`1

---