syssvm: Run sysSVM

In thmourikis/sysSVM: sysSVM

Description

syssvm uses e1071 to run an one-class SVM

Usage

syssvm(input.file = "example_data/oac_ML_input_formatted.tsv",
  output.dir = "OAC", exclude.features = c("young", "no_ALL_muts"),
  models = NULL, scaling.factors = NULL, cv = 3, iters = 1000,
  step = 100, top.rank = 10, kernels = c("linear", "polynomial", "radial",
  "sigmoid"), cores = 2, ncg.tissue.name = "esophagus", refine = TRUE)

Arguments

input.file	Path to input file [default is "example_data/oac_ML_input_formatted.tsv"]
output.dir	Path to output directory where sysSVM results will be written [default is "OAC"]
exclude.features	Features to be excluded from training/prediction [default is c("young", "no_ALL_muts")]
models	Functionality to use already trained models for new samples
scaling.factors	Object with sysSVM scaling factors. Use this to make predictions using already-trained sysSVM models
cv	k-fold cross-validation [default is 3]
iters	Iterations to be performed during ross-validation [default is 1000]
step	Step to summarise results (i.e. every step iterations) [default is 100]
top.rank	Sys-candidates to be considered per patient [default is 10]
kernels	Kernels to be used for training/prediction [default is c("linear", "polynomial", "radial", "sigmoid")]
cores	Multi-core functionality [default is 2]
ncg.tissue.name	NCG tissue for refinement of sys-candidates [default is "esophagus"]
refine	set to TRUE to refine scores based on gtex expression and previously published results

Examples

syssvm()

thmourikis/sysSVM documentation built on May 30, 2019, 4:05 a.m.