Function reference • knockofftools

All functions

eval_fdp(): False discovery proportion (fdp) as function of selection and known negatives:

eval_tpp(): True positive proportion (tpp) as function of selection and known positives:

generate_X(): Simulate Gaussian and binary covariate predictors

generate_simdata(): Generate simulated data set for vignettes

generate_y(): Simulate Gaussian response from a sparse regression model

glasso_adjacency_matrix(): Estimate adjacency matrix using graphical LASSO (glasso)

knockoff(): Knockoffs for general covariate data frames

knockoff.statistics(): Knockoff (feature) statistics:

knockoffs_mx(): Gaussian MX-knockoffs for continuous variables

knockoffs_seq(): Sequential knockoffs for continuous and categorical variables

knockoffs_sparse_seq(): Sparse sequential knockoff generation algorithm

multi_select(): Select variables based on the heuristic multiple selection algorithm from Kormaksson et al. 'Sequential knockoffs for continuous and categorical predictors: With application to a large psoriatic arthritis clinical trial pool.' Statistics in Medicine. 2021;1–16.

plot(<variable.selections>): Heatmap of multiple variable selections ordered by importance

selections_control_FDR(): Controls the false discovery rate (FDR) given knockoff W-statistics.

selections_control_PFER(): Controls the per-familywise error rate (PFER) given knockoff W-statistics.

selections_control_kFWER(): Controls the k-familywise error rate (k-FWER) given a vector of knockoff W-statistics.

sim_glmnet(): Simulate from glmnet penalized regression model

sim_simple(): Simple knockoff generator based on least squares fit (continuous variables) or multinomial logistic regression (factor variables) respectively. If X is empty, knockoffs are sampled from the marginal distribution of y

simulWeib(): Function that simulates response from Cox model with Weibull baseline hazard:

stat_glmnet(): Knockoff (feature) statistics: Absolute elastic-net coefficient differences between original and knockoff variables

stat_predictive_causal_forest(): Causal forest based knockoff (feature) statistics that captues the predictive strength: Difference from importance scores derived by causal forest

stat_predictive_glmnet(): Knockoff (feature) statistics that captues the predictive strength: Absolute coefficient differences between treatment original variables interaction terms and treatment knockoff variables interaction terms

stat_random_forest(): Knockoff (feature) statistics: Random forest

variable.selections(): Knockoff variable selection: Select the variables by controlling a user-specified error rate