admm_algorithm |
Performs the Alternating Descent Method of Multipliers (ADMM) algorithm to estimate the LoRI parameters. |
aravo |
Aravo |
default_projection |
Default projection on covariate space: center by row and column. |
estimate_null_model |
Estimates the LoRI model under the constraint Theta = 0. |
gradient |
Computes the gradient of the objective function in X for gradient descent step in ADMM. |
lambda_cv |
Selects the parameter $lambda_CV$ with cross-validation. |
lambda_QUT |
Computes the threshold $lambda_QUT$ with parametric bootstrap when NO covariates are available. If you don't have any covariates, use this function instead of 'lambda_QUT_covariates' which will be significantly slower. |
lambda_QUT_covariates |
Computes the threshold $lambda_QUT$ with parametric bootstrap when covariates are available. If you don't have any covariates, use the function 'lambda_QUT' which will be significantly faster. |
lori |
Main function to be used to fit the LORI model |
objective_function |
Computes the value of the objective function in X to be optimized at each iteration of ADMM. |
plot_interaction |
Plot the rows and columns of the contingency table in the Euclidean space defined by two of the first principal directions of the interaction matrix Theta. Theta corresponds to the interaction remaining after discarding the effects of the covariates. The interpretation is the following. A row and a column that are close in Euclidean distance interact highly. Two rows or two columns that are close in Euclidean distance have similar profiles. |