generate_synth_data: Generate 'multi-omic' synthetic data matrices for validation
In agapow/subtypr: Robust and Validated Patient Subtyping for Precision Medicine

Description Usage Arguments Details Value

For validation and testing multi-omic / integrative methods, we provide two type of synthetic data:

Unstructured data with no cluster. The idea is to check the results of a method on data with no real structure.
Artificially structured data based on a real dataset.

1
2
3

generate_synth_data(type = c("gaussian", "uniform", "moclus"),
  n_samples = NULL, n_features = NULL, n_layers = NULL,
  support_data_list = NULL, ...)

`type`	The type of data. See Details.
`n_samples`	The number of samples of the generated data.
`n_features`	A vector of integer: the numbers of features for each generated matrix.
`n_layers`	integer. The number of layers (features matrix) to be generated.
`support_data_list`	A list of features matrix that are used to generate the structured data.
`...`	parameters to be passed to `generate_multi_structured()`. See ?generate_multi_structured.

For unstructured type, the dimensions of the data are needed. (n_samples, n_features and n_layers). See ?generate_multi_unstructured. For structured type, a list of features matrix, support_data_list is needed as a support to generate the data. There is also special parameters to be set, see ?generate_multi_structured for more information.

'multi-omic' synthetic data matrices for validation.

agapow/subtypr documentation built on May 5, 2019, 1:33 a.m.