Predictions of a ppLasso or gr_ppLasso object

Return the model predictions of a ppLasso or gr_ppLasso object

Usage

# S3 method for class 'ppLasso'
predict(
  fit,
  data,
  Z.char,
  prov.char,
  lambda,
  which = 1:length(fit$lambda),
  type = c("link", "response", "class", "vars", "nvars"),
  ...
)

# S3 method for class 'gr_ppLasso'
predict(
  fit,
  data,
  Z.char,
  prov.char,
  lambda,
  which = 1:length(fit$lambda),
  type = c("link", "response", "class", "vars", "groups", "nvars", "ngroups",
    "beta.norm"),
  ...
)

Arguments

fit

a ppLasso or gr_ppLasso.

data

an dataframe or list object that contains the variables for prediction.

Z.char

names of covariates in data as vector of character strings.

prov.char

name of provider IDs variable in data as a character string.

lambda

values of the regularization parameter lambda at which predictions are requested. For values of lambda not in the sequence of fitted models, linear interpolation is used.

which

indices of the penalty parameter lambda at which predictions are required. By default, all indices are returned. If lambda is specified, this will override which.

type

type of prediction:

• link: linear predictors

• response: fitted values (i.e., exp(link)/(1 + exp(link)))

• class: the binomial outcome with the largest probability

• vars: the indices for the non-zero coefficients

• nvars: the number of non-zero coefficients

• groups: the indices for the non-zero groups

• ngroups: the number of non-zero coefficients

• beta.norm: L2 norm of the coefficients in each group

...

Examples

data(BinaryData)
data <- BinaryData$data
Y.char <- BinaryData$Y.char
prov.char <- BinaryData$prov.char
Z.char <- BinaryData$Z.char
fit <- pp.lasso(data, Y.char, Z.char, prov.char)
predict(fit, data, Z.char, prov.char, lambda = fit$lambda, type = "response")[1:10, 1:5]
#>           0.132    0.1202    0.1096    0.0998     0.091
#>  [1,] 0.2874999 0.2996927 0.3106971 0.3206762 0.3297643
#>  [2,] 0.2051302 0.2112109 0.2163924 0.2208163 0.2246015
#>  [3,] 0.2051302 0.2191951 0.2321638 0.2441339 0.2551989
#>  [4,] 0.3947391 0.4208778 0.4449423 0.4671087 0.4875485
#>  [5,] 0.2345694 0.2340619 0.2330311 0.2316163 0.2299295
#>  [6,] 0.3188411 0.3744516 0.4280555 0.4786478 0.5256182
#>  [7,] 0.3924033 0.3836025 0.3752418 0.3673071 0.3597819
#>  [8,] 0.3658512 0.3870533 0.4065558 0.4245437 0.4411742
#>  [9,] 0.2298886 0.2256943 0.2215601 0.2175107 0.2135686
#> [10,] 0.3947391 0.4374814 0.4772352 0.5139266 0.5476058
predict(fit, data, Z.char, prov.char, lambda = 0.001, type = "class")[1:10]
#>  [1] 1 0 0 1 0 1 0 1 0 1
predict(fit, data, Z.char, prov.char, lambda = 0.04, type = "vars")
#> Z1 Z3 Z4 
#>  1  3  4 

data(BinaryData)
data <- BinaryData$data
Y.char <- BinaryData$Y.char
prov.char <- BinaryData$prov.char
Z.char <- BinaryData$Z.char
group <- BinaryData$group
fit <- grp.lasso(data, Y.char, Z.char, prov.char, group = group)
predict(fit, data, Z.char, prov.char, lambda = fit$lambda, type = "response")[1:10, 1:5]
#>          0.0939    0.0856     0.078     0.071    0.0647
#>  [1,] 0.2874998 0.2986935 0.3093832 0.3193357 0.3255634
#>  [2,] 0.2051288 0.2100619 0.2149546 0.2191076 0.2136753
#>  [3,] 0.2051288 0.2180235 0.2305894 0.2431181 0.2603558
#>  [4,] 0.3947376 0.4185762 0.4419123 0.4646744 0.4888233
#>  [5,] 0.2345683 0.2353836 0.2346598 0.2323132 0.2223927
#>  [6,] 0.3188407 0.3711201 0.4233544 0.4745208 0.5275946
#>  [7,] 0.3924044 0.3852951 0.3773617 0.3679648 0.3448889
#>  [8,] 0.3658529 0.3891889 0.4090002 0.4266957 0.4552054
#>  [9,] 0.2298861 0.2270617 0.2232482 0.2183580 0.2074687
#> [10,] 0.3947376 0.4344220 0.4731255 0.5103060 0.5458409
predict(fit, data, Z.char, prov.char, lambda = 0.001, type = "class")[1:10]
#>  [1] 1 0 0 1 0 1 0 1 0 1
predict(fit, data, Z.char, prov.char, lambda = 0.04, type = "vars")
#> Z1 Z2 Z3 Z4 
#>  1  2  3  4 
predict(fit, data, Z.char, prov.char, lambda = 0.04, type = "groups")
#> [1] 1 2
#> Levels: 1 2 3