Articles • metasnf

Getting Started: An introduction to the metasnf package and installation instructions.

The Settings Matrix: The object that controls all hyperparameters defining the space of cluster solutions to explore.
The Data List: The main object used to store data in the metasnf package.

SNF Schemes: Controlling the way that individual input dataframes are combined into a final fused network.
Distance Metrics: Vary distance metrics to expand or refine the space of generated cluster solutions.
Clustering Algorithms: Vary clustering algorithm to expand or refine the space of generated cluster solutions.
Feature Weighting: Vary feature weights to expand or refine the space of generated cluster solutions.

Stability Measures and Consensus Clustering: Evaluating robustness of cluster solutions through resampling methods.
Quality Measures: Calculate context-agnostic measures of clustering compactness and separation.
Unwanted Signal: Linearly regress out unwanted signal from variables for clustering.
Parallel Processing: Leverage parallel processing to speed up the metaSNF pipeline.
Label Propagation: Validate or extend cluster insights to new observations through semi-supervised label propagation.
Imputations: Incorporate imputation approach as another source of variability in the generated space of cluster solutions.
NMI Scores: Calculate how important various features were to the final SNF cluster solution.

Correlation Plots: Visualize correlations between data prior to clustering.
Similarity Matrices: Visualize the affinity matrices produced by SNF and how they associate with other data attributes.
Manhattan Plots: Visualize a summary of the association between cluster-feature and feature-feature relationships.
Feature Plots: Visualize how features are distributed within a cluster solution.
Alluvial Plots: Visualize how cluster number influences the distribution of observations.