Data Set-up | Pre-processing | Generating the data list | Defining sets of hyperparameters to use for SNF and clustering | The settings data frame | Other parts of the SNF config | Running SNF and clustering | Identifying and visualizing meta clusters | Characterizing cluster solutions | Calculating associations between cluster solutions and initial data | Visualizing feature associations with meta clustering results | Characterizing individual solutions representative of each meta cluster | Relating results to metasnf hyperparameters | Quality measures | Stability measures | Evaluating separation across "target features" of importance | Validating results with label propagation | References

Default clustering | Other built-in clustering options | Structure of a clustering algorithm function | Non-automated clustering | Example of non-automated clustering: DBSCAN

Data set-up

Creating a default SNF config | The settings data frame | The distance functions list | The clustering functions list | The weights matrix | Customizing an SNF config | Alpha, k, and t | Inclusion columns and data frame dropout | Grid searching | Assembling an SNF config in pieces | "settings_df building failed to converge"

The original SNF example | 1. Load the package | 2. Set SNF hyperparameters | 3. Load the data | 4. Generate similarity matrices for each data source | 5. Integrate similarity matrices with SNF | 6. Find clusters in the integrated matrix | The same example using metasnf | 2. Store the data in a data list | 3. Store all the settings of the desired SNF runs in an SNF config | 4. Run SNF | References

Accounting for confounding features | Unwanted signal | Procedure using the metasnf package | Limitations and important considerations | 1. Excessive loss of signal | 2. Lack of accounting for non-linearities | 3. Inability to adjust ordinal, discrete, or categorical data

Data set-up | Heatmaps

Distance functions | How the dist_fns_list is used | Removing the default distance_metrics | Supplying weights to distance metrics | Custom distance metrics | Requesting metrics | List of prewritten distance metrics functions | References

Generating and Using the Weights Matrix | The Nitty Gritty of How Weights are Used

Introduction | Installation

Data set-up | Associations with Multiple Cluster Solutions (esm_manhattan_plot) | Associations with Meta Clusters (mc_manhattan_plot) | Associations with a Key Feature

Basic usage | Including a progress bar | Number of processes

Data set-up | Visualize similarity matrices sorted by cluster label | Annotations | More on sorting

(1) "Individual" | (2) "Two-step" | (3) "Domain" | Custom SNF schemes

The data_list