fit_inspre_sequence | R Documentation |
Fits inspre model for sequence of lambda values.
fit_inspre_sequence(
X,
lambda,
W = NULL,
rho = 1,
its = 100,
delta_target = 1e-04,
symmetrize = FALSE,
verbose = 1,
gamma = 0,
train_prop = 1,
mu = 5,
tau = 2.5,
solve_its = 3,
ncores = 1,
warm_start = FALSE,
constraint = "UV",
DAG = FALSE
)
X |
DxD Matrix to find approximate sparse inverse of. |
lambda |
Float or sequence of floats. Path of L1 regularization strength on inverse of X. |
W |
DxD Matrix of weights. |
rho |
Float. Initial learning rate for ADMM. |
its |
Integer. Maximum number of iterations. |
delta_target |
Float. Target change in solution. |
symmetrize |
True to force the output to be symmetric. If the input is symmetric, the output isn't always perfectly symmetric due to numerical issues. |
verbose |
0, 1 or 2. 2 to print convergence progress for each lambda, 1 to print convergence result for each lambda, 0 for no output. |
gamma |
Float or sequence of floats. Determinant penalty strength. |
train_prop |
Proportion of the dataset to use to train the model. Set to 1.0 to use all data. |
mu |
rho modification parameter for ADMM. Rho will be increased/decreased when the dual constrant and primal constraint are off by a factor of > mu. |
tau |
rho modification parameter for ADMM. When called for, rho will be increased/decreased by the factor tau. |
solve_its |
Integer, number of iterations of bicgstab/lasso to run for each U and V update. |
ncores |
Integer, number of cores to use. |
warm_start |
Boolean, TRUE to start next fit with result of previous fit. Default FALSE. |
constraint |
One of "UV" or "VU". Constraint to use. |
DAG |
Bool. True to resitrict solutions to approximate DAGs. |
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.