Package 'xtifedml' reference manual

Title:	Double Machine Learning for Static Panel Models with Interactive Fixed Effects
Description:	Implements partially linear panel regression (PLPR) models with interactive fixed effects, high-dimensional confounding variables, and an exogenous treatment variable within the double machine learning framework. Estimates the structural parameter (treatment effect) in static panel data models with interactive fixed effects using the approach established in Chen et al. (2026) <doi:10.48550/arXiv.2608.01137>. Builds on the object-oriented package 'DoubleML' (Bach et al., 2024) <doi:10.18637/jss.v108.i03> and 'xtdml' (Polselli, 2025) <doi:10.48550/arXiv.2512.15965>, using the 'mlr3' ecosystem.
Authors:	Binzhi Chen [aut] (ORCID: <https://orcid.org/0000-0002-5094-7740>), Annalivia Polselli [aut, cre] (ORCID: <https://orcid.org/0009-0002-7579-7926>)
Maintainer:	Annalivia Polselli <[email protected]>
License:	GPL-2 \| GPL-3
Version:	0.1.4
Built:	2026-09-30 14:13:53 UTC
Source:	https://github.com/cran/xtifedml

Data generating process for partially linear panel regression models with interactive fixed effects

Description

Generates data from a partially linear regression model for panel data with fixed effects from Binzhi et al. (2026).

The data generating process is defined as

$Y_{it} = \theta D_{it} + l_0(X_{it}) + \lambda_i'f_t + U_{it},$ $D_{it} = m_0(X_{it}) + \lambda_{i}'f_{t} + V_{it},$

where $U_{it} \sim \mathcal{N}(0,1)$ , $V_{it} \sim \mathcal{N}(0,1)$ , $\lambda_i\sim \mathcal{N}(0,2)$ , $f_t\sim \mathcal{N}(0,2)$ .

The covariates are distributed as $X_{it,p} \sim \mathcal{N}(1, 5) + \lambda_i'f_t$ , where $p$ is the number of covariates.

The nuisance functions are generated as follows. For dgp = "linear",

$m_0(X_{it}) = a_1 X_{it,1} + a_2 X_{it,3},$

$g_0(X_{it}) = b_1 X_{it,1} + b_2 X_{it,3} .$

For dgp = "smooth",

$m_0(X_{it}) = a_1 max\{X_{it,1},0\} + a_2 |X_{it,3}|,$

$g_0(X_{it}) = b_1 [X_{it,1} \times 1\{X_{it,3}>0\}] + b_2 [X_{it,3} \times X_{it,3}>0),$

For dgp = "discouNTinuous",

$m_0(X_{it}) = a_1 [X_{it,1} \times 1(X_{it,1}>0)] + a_2 [X_{it,1} \times X_{it,3}],$

$g_0(X_{it}) = b_1 [X_{it,1} \times X_{it,3}] + b_2 [X_{it,3} \times 1(X_{it,3}>0)].$

where $a_1=b_2=0.25$ and $a_2=b_1=0.5$ .

Usage

make_plpr_data(
  n_obs = 500,
  t_per = 10,
  dim_x = 20,
  theta = 1,
  r = 2,
  rho = 0.6,
  dgp = NULL
)
make_plpr_data(
  n_obs = 500,
  t_per = 10,
  dim_x = 20,
  theta = 1,
  r = 2,
  rho = 0.6,
  dgp = NULL
)

Arguments

n_obs

(integer(1))
The number of subjects (cross-sectional units) to simulate. Default 500.

t_per

(integer(1))
The number of time periods to simulate. Default 10.

dim_x

(integer(1))
The number of covariates. Default 20.

theta

(numeric(1))
The value of the causal parameter. Default 1.

r

(numeric(1))
Number of factors. Default 2.

rho

(numeric(1))
Parameter governing the correlation between the IFE and the included variables. Default 0.6.

dgp

(character(1))
The type of dataset to generate. Default NULL.

Value

A data object.

Examples

df = make_plpr_data(n_obs = 500, t_per = 10, dim_x = 20, theta = 1, r=2, dgp = "discontinuous")

df = make_plpr_data(n_obs = 500, t_per = 10, dim_x = 20, theta = 1, r=2, dgp = "discontinuous")

Abstract Class xtifedml

Description

Abstract base class that cannot be initialized directly.

Implements partially linear panel regression (PLPR) models with high-dimensional confounding variables and an exogenous treatment variable within the double machine learning framework. Estimates the structural parameter (treatment effect) in static panel data models with interactive fixed effects, using the approach of Binzhi et al. (2026).

Builds on the object-oriented architecture of DoubleML (Bach et al., 2024) and xtdml (Polselli, 2025), using the 'mlr3' ecosystem and the 'R6' package.

Format

R6::R6Class object.

Active bindings

all_coef_theta: (matrix())
Estimates of the causal parameter(s) "theta" for the n_rep different sample splits after calling fit().
all_dml1_coef_theta: (array())
Estimates of the causal parameter(s) "theta" for the n_rep different sample splits after calling fit() with dml_procedure = "dml1".
all_se_theta: (matrix())
Standard errors of the causal parameter(s) "theta" for the n_rep different sample splits after calling fit().
all_model_rmse: (matrix())
Model root-mean-squared-error.
apply_cross_fitting: (logical(1))
Indicates whether cross-fitting should be applied. Default is TRUE.
coef_theta: (numeric())
Estimates for the causal parameter(s) "theta" after calling fit().
data: (data.table)
Data object.
dml_procedure: (character(1))
A character() ("dml1" or "dml2") specifying the double machine learning algorithm. Default is "dml2".
draw_sample_splitting: (logical(1))
Indicates whether the sample splitting should be drawn during initialization of the object. Default is TRUE.
learner: (named list())
The machine learners for the nuisance functions.
n_folds: (integer(1))
Number of folds. Default is 5.
n_rep: (integer(1))
Number of repetitions for the sample splitting. Default is 1.
params: (named list())
The hyperparameters of the learners.
psi_theta: (array())
Value of the score function $\psi(W;\theta_0,\eta_0)=-\psi_a(W;\eta_0) \theta_0 + \psi_b(W;\eta_0)$ after calling fit().
psi_theta_a: (array())
Value of the score function component $\psi_a(W;\eta_0)$ after calling fit().
psi_theta_b: (array())
Value of the score function component $\psi_b(W;\eta_0)$ after calling fit().
res_y: (array())
Residual of output equation
res_d: (array())
Residual of treatment equation
predictions: (array())
Predictions of the nuisance models after calling fit(store_predictions=TRUE).
targets: (array())
Targets of the nuisance models after calling fit(store_predictions=TRUE).
rmses: (array())
The root-mean-squared-errors of the nuisance parameters
all_model_mse: (array())
Collection of all mean-squared-errors of the model
model_rmse: (array())
The root-mean-squared-errors of the model
models: (array())
The fitted nuisance models after calling fit(store_models=TRUE).
pval_theta: (numeric())
p-values for the causal parameter(s) "theta" after calling fit().
score: (character(1))
A character(1) specifying the score function among "orth-PO", "orth-IV". Default is "orth-PO".
se_theta: (numeric())
Standard errors for the causal parameter(s) "theta" after calling fit().
smpls: (list())
The partition used for cross-fitting.
smpls_cluster: (list())
The partition used for cross-fitting. smpl is at cluster-var
t_stat_theta: (numeric())
t-statistics for the causal parameter(s) "theta" after calling fit().
tuning_res_theta: (named list())
Results from hyperparameter tuning.

Methods

`xtifedml$new()`

DML with IFE is an abstract class that can't be initialized.

Usage

xtifedml$new()

`xtifedml$print()`

Print 'DML with IFE' objects.

Usage

xtifedml$print()

`xtifedml$fit()`

Estimate DML models with IFE.

Usage

xtifedml$fit(store_predictions = FALSE, store_models = FALSE)

Arguments

store_predictions: (logical(1))
Indicates whether the predictions for the nuisance functions should be stored in field predictions. Default is FALSE.
store_models: (logical(1))
Indicates whether the fitted models for the nuisance functions should be stored in field models if you want to analyze the models or extract information like variable importance. Default is FALSE.

Returns

self

`xtifedml$split_samples()`

Draw sample splitting for Double ML models with IFE.

The samples are drawn according to the attributes n_folds, n_rep and apply_cross_fitting.

Usage

xtifedml$split_samples()

Returns

self

`xtifedml$tune()`

Hyperparameter-tuning for DML models with IFE.

The hyperparameter-tuning is performed using the tuning methods provided in the mlr3tuning package. For more information on tuning in mlr3, we refer to the section on parameter tuning in the mlr3 book.

Usage

xtifedml$tune(
  param_set,
  tune_settings = list(n_folds_tune = 5, rsmp_tune = mlr3::rsmp("cv", folds = 5), measure
    = NULL, terminator = mlr3tuning::trm("evals", n_evals = 20), tuner =
    mlr3tuning::tnr("grid_search", resolution = 5)),
  tune_on_folds = FALSE
)

Arguments

param_set

(named list())
A named list with a parameter grid for each nuisance model/learner (see method learner_names()). The parameter grid must be an object of class ParamSet.

tune_settings

(named list())
A named list() with arguments passed to the hyperparameter-tuning with mlr3tuning to set up TuningInstance objects. tune_settings has entries

terminator (Terminator)
A Terminator object. Specification of terminator is required to perform tuning.
tuner (Tuner)
A Tuner object created with tnr(), which defines the optimization algorithm (e.g., tnr("grid_search", resolution = 10) or tnr("random_search")). Any tuner-specific arguments, such as resolution for "grid_search", must be supplied directly to tnr() when constructing this object. If not specified by the user, default is tnr("grid_search", resolution = 5).
rsmp_tune (Resampling or character(1))
A Resampling object (recommended) or option passed to rsmp() to initialize a Resampling for parameter tuning in mlr3. If not specified by the user, default is set to "cv" (cross-validation).
n_folds_tune (integer(1), optional)
If rsmp_tune = "cv", number of folds used for cross-validation. If not specified by the user, default is set to 5.
measure (NULL, named list(), optional)
Named list containing the measures used for parameter tuning. Entries in list must either be Measure objects or keys to be passed to passed to msr(). The names of the entries must match the learner names (see method learner_names()). If set to NULL, default measures are used, i.e., "regr.mse" for continuous outcome variables and "classif.ce" for binary outcomes.

tune_on_folds

(logical(1))
Indicates whether the tuning should be done fold-specific or globally. Default is FALSE.

Returns

self

`xtifedml$summary()`

Summary for DML models with IFE after calling fit().

Usage

xtifedml$summary(digits = max(3L, getOption("digits") - 3L))

Arguments

digits: (integer(1))
The number of significant digits to use when printing.

`xtifedml$confint()`

Confidence intervals for DML models with IFE.

Usage

xtifedml$confint(parm, joint = FALSE, level = 0.95)

Arguments

parm: (numeric() or character())
A specification of which parameters are to be given confidence intervals among the variables for which inference was done, either a vector of numbers or a vector of names. If missing, all parameters are considered (default).
joint: (logical(1))
Indicates whether joint confidence intervals are computed. Default is FALSE.
level: (numeric(1))
The confidence level. Default is 0.95.

Returns

A matrix() with the confidence interval(s).

`xtifedml$learner_names()`

Returns the names of the learners.

Usage

xtifedml$learner_names()

Returns

character() with names of learners.

`xtifedml$params_names()`

Returns the names of the nuisance models with hyperparameters.

Usage

xtifedml$params_names()

Returns

character() with names of nuisance models with hyperparameters.

`xtifedml$set_ml_nuisance_params()`

Set hyperparameters for the nuisance models of DML models with IFE.

Note that in the current implementation, either all parameters have to be set globally or all parameters have to be provided fold-specific.

Usage

xtifedml$set_ml_nuisance_params(
  learner = NULL,
  treat_var = NULL,
  params,
  set_fold_specific = FALSE
)

Arguments

learner: (character(1))
The nuisance model/learner (see method params_names).
treat_var: (character(1))
The treatment variAble (hyperparameters can be set treatment-variable specific).
params: (named list())
A named list() with estimator parameters for time-varying covariates. Parameters are used for all folds by default. Alternatively, parameters can be passed in a fold-specific way if option fold_specificis TRUE. In this case, the outer list needs to be of length n_rep and the inner list of length n_folds_per_cluster.
set_fold_specific: (logical(1))
Indicates if the parameters passed in params should be passed in fold-specific way. Default is FALSE. If TRUE, the outer list needs to be of length n_rep and the inner list of length n_folds_per_cluster. Note that in the current implementation, either all parameters have to be set globally or all parameters have to be provided fold-specific.

Returns

self

`xtifedml$get_params()`

Get hyper-parameters for the nuisance model of xtifedml models.

Usage

xtifedml$get_params(learner)

Arguments

learner: (character(1))
The nuisance model/learner (see method params_names())

Returns

named list()with paramers for the nuisance model/learner.

`xtifedml$clone()`

The objects of this class are cloneable with this method.

Usage

xtifedml$clone(deep = FALSE)

Arguments

deep: Whether to make a deep clone.

Data Setup for Panel Data Approaches with Cluster Variables

Description

Constructs a double machine learning (DML) data backend for panel data, supporting designs with one or two cluster variables.

xtifedml_data objects can be initialized from a data.table. The following functions can be used to create a new instance of xtifedml_data.

xtifedml_data$new() for initialization from a data.table.
xtifedml_data_df() for initialization from a data.frame.

Active bindings

all_variables: (character())
All variables available in the data frame.
d_cols: (character())
The treatment variable.
dbar_col: (NULL, character()')
The individual mean of the treatment variable.
data: (data.table)
Data object.
data_model: (data.table)
Internal data object that implements the causal panel model as specified by the user via y_col, d_cols, x_cols, dbar_col.
n_obs: (integer(1))
The number of observations.
n_treat: (integer(1))
The number of treatment variables.
treat_col: (character(1))
"Active" treatment variable in the multiple-treatment case.
x_cols: (character())
The covariates.
y_col: (character(1))
The outcome variable.
panel_id: (character())
The panel identifier.
time_id: (character())
The time identifier.
cluster_cols: (character())
The cluster variable(s).
n_cluster_vars: (integer(1))
The number of cluster variables.
approach: (character(1))
A character() ("fd-exact", "wg-approx" or "cre") specifying the panel data technique to apply to estimate the causal model. Default is "fd-exact".
transformX: (character(1))
A character() ("no", "minmax" or "poly") specifying the type of transformation to apply to the X data. "no" does not transform the covariates X and is recommended for tree-based learners. "minmax" applies the Min-Max normalization $x' = (x-x_{min})/(x_{max}-x_{min})$ to the covariates and is recommended with neural networks. "poly" add polynomials up to order three and interactions between all possible combinations of two and three variables; this is recommended for Lasso. Default is "no".

Methods

Public methods

xtifedml_data$new()
xtifedml_data$print()
xtifedml_data$set_data_model()
xtifedml_data$clone()

`xtifedml_data$new()`

Creates a new instance of this R6 class.

Usage

xtifedml_data$new(
  data = NULL,
  x_cols = NULL,
  y_col = NULL,
  d_cols = NULL,
  dbar_col = NULL,
  panel_id = NULL,
  time_id = NULL,
  cluster_cols = NULL,
  approach = NULL,
  transformX = NULL
)

Arguments

data: (data.table, data.frame())
Data object.
x_cols: (character())
y_col: (character(1))
The outcome variable.
d_cols: (character(1))
The treatment variable.
dbar_col: (NULL, character()⁠) \cr Individual mean of the treatment variable (used for the CRE approach). Default is ⁠NULL'.
panel_id: (character())
The panel identifier.
time_id: (character())
The time identifier.
cluster_cols: (character())
The cluster variable(s).
approach: (character(1))
A character() ("fd-exact", "wg-approx" or "cre") specifying the panel data technique to apply to estimate the causal model. Default is "fd-exact".
transformX: (character(1))
A character() ("no", "minmax" or "poly") specifying the type of transformation to apply to the X data. "no" does not transform the covariates X and is recommended for tree-based learners. "minmax" applies the Min-Max normalization $x' = (x-x_{min})/(x_{max}-x_{min})$ to the covariates and is recommended with neural networks. "poly" add polynomials up to order three and interactions between all possible combinations of two and three variables; this is recommended for Lasso. Default is "no".

`xtifedml_data$print()`

Print xtifedml_data objects.

Usage

xtifedml_data$print()

`xtifedml_data$set_data_model()`

Setter function for data_model. The function implements the causal model as specified by the user via y_col, d_cols, x_cols, panel_id, time_id and cluster_cols and assigns the role for the treatment variables in the multiple-treatment case.

Usage

xtifedml_data$set_data_model(treatment_var)

Arguments

treatment_var: (character())
Active treatment variable that will be set to treat_col.

`xtifedml_data$clone()`

The objects of this class are cloneable with this method.

Usage

xtifedml_data$clone(deep = FALSE)

Arguments

deep: Whether to make a deep clone.

Wrapper for Double machine learning data-backend initialization from data.frame.

Description

Initalization of DoubleMLData from data.frame.

Usage

xtifedml_data_df(
  df,
  x_cols = NULL,
  y_col = NULL,
  d_cols = NULL,
  panel_id = NULL,
  time_id = NULL,
  cluster_cols = NULL,
  approach = NULL,
  transformX = NULL
)
xtifedml_data_df(
  df,
  x_cols = NULL,
  y_col = NULL,
  d_cols = NULL,
  panel_id = NULL,
  time_id = NULL,
  cluster_cols = NULL,
  approach = NULL,
  transformX = NULL
)

Arguments

df

(data.frame())
Data object.

x_cols

(character())
The covariates.

y_col

(character(1))
The outcome variable.

d_cols

(character())
The treatment variable(s).

panel_id

(NULL, character())
The panel identifier. Default is NULL.

time_id

(NULL, character())
The time identifier. Default is NULL.

cluster_cols

(NULL, character())
The cluster variables. Default is panel_id.

approach

(character(1))
A character() ("pooled","pcce") specifying the panel data technique to apply to estimate the causal model. Default is "pcce".

transformX

(character(1))
A character() ("no", "minmax" or "poly") specifying the type of transformation to apply to the X data. "no" does not transform the covariates X and is recommended for tree-based learners. "minmax" applies the Min-Max normalization $x' = (x-x_{min})/(x_{max}-x_{min})$ to the covariates and is recommended with neural networks. "poly" add polynomials up to order three and interactions between all possible combinations of two and three variables; this is recommended for Lasso. Default is "no".

Value

Creates a new instance of class xtifedml_data.

Examples


# Generate simulated panel dataset from `xtifedml`
data = make_plpr_data(n_obs = 500, t_per = 10, dim_x = 20, theta = 1, r=2, dgp = "discontinuous")

# Set up DML data environment
x_cols  = paste0("X", 1:20)

obj_xtifedml_data = xtifedml_data_df(data,
                x_cols = x_cols,  y_col = "y", d_cols = "d",
                panel_id = "id",
                time_id = "time",
                cluster_cols = "id",
                approach = "pcce",
                transformX = "no")

obj_xtifedml_data$print()

# Generate simulated panel dataset from `xtifedml`
data = make_plpr_data(n_obs = 500, t_per = 10, dim_x = 20, theta = 1, r=2, dgp = "discontinuous")

# Set up DML data environment
x_cols  = paste0("X", 1:20)

obj_xtifedml_data = xtifedml_data_df(data,
                x_cols = x_cols,  y_col = "y", d_cols = "d",
                panel_id = "id",
                time_id = "time",
                cluster_cols = "id",
                approach = "pcce",
                transformX = "no")

obj_xtifedml_data$print()

Routine to estimate partially linear panel regression models with fixed effects within double machine learning.

Description

Routine to estimate partially linear panel regression models with fixed effects within double machine learning.

Format

R6::R6Class object inheriting from xtifedml.

Details

Consider partially linear panel regression (PLR) model of form

$Y_{it} = \theta D_{it} + l_0(X_{it}) + \lambda_i'f_t + U_{it},$ $D_{it} = m_0(X_{it}) + \lambda_{i}'f_{t} + V_{it}.$

Super class

xtifedml -> xtifedml_plr

Methods

Public methods

xtifedml_plr$new()
xtifedml_plr$set_ml_nuisance_params()
xtifedml_plr$tune()
xtifedml_plr$clone()

Inherited methods

xtifedml$confint()
xtifedml$fit()
xtifedml$get_params()
xtifedml$learner_names()
xtifedml$params_names()
xtifedml$print()
xtifedml$split_samples()
xtifedml$summary()

`xtifedml_plr$new()`

Creates a new instance of this R6 class.

Usage

xtifedml_plr$new(
  data,
  ml_l,
  ml_m,
  ml_g = NULL,
  n_folds = 5,
  n_rep = 1,
  score = "orth-PO",
  dml_procedure = "dml2",
  draw_sample_splitting = TRUE,
  apply_cross_fitting = TRUE
)

Arguments

data: (xtifedml_data)
The xtifedml_data object providing the data and specifying the variables of the causal model.
ml_l: (LearnerRegr, Learner, character(1))
A learner of the class LearnerRegr, which is available from mlr3 or its extension packages mlr3learners or mlr3extralearners. Alternatively, a Learner object with public field task_type = "regr" can be passed, for example of class GraphLearner. The learner can possibly be passed with specified parameters, for example lrn("regr.cv_glmnet", s = "lambda.min").
ml_l refers to the nuisance function $l_0(X) = E[Y|X]$ .
ml_m: (LearnerRegr, LearnerClassif, Learner, character(1))
A learner of the class LearnerRegr, which is available from mlr3 or its extension packages mlr3learners or mlr3extralearners. For binary treatment variables, an object of the class LearnerClassif can be passed, for example lrn("classif.cv_glmnet", s = "lambda.min"). Alternatively, a Learner object with public field task_type = "regr" or task_type = "classif" can be passed, respectively, for example of class GraphLearner.
ml_m refers to the nuisance function $m_0(X) = E[D|X]$ .
ml_g: (LearnerRegr, Learner, character(1))
A learner of the class LearnerRegr, which is available from mlr3 or its extension packages mlr3learners or mlr3extralearners. Alternatively, a Learner object with public field task_type = "regr" can be passed, for example of class GraphLearner. The learner can possibly be passed with specified parameters, for example lrn("regr.cv_glmnet", s = "lambda.min").
ml_g refers to the nuisance function $g_0(X) = E[Y - D\theta_0|X]$ . Note: The learner ml_g is only required for the score 'IV-type'. Optionally, it can be specified and estimated for callable scores.
n_folds: (integer(1))
Number of folds. Default is 5.
n_rep: (integer(1))
Number of repetitions for the sample splitting. Default is 1.
score: (character(1))
A character(1) ("orth-PO" or "orth-IV"). "orth-PO" is Neyman-orthogonal score with the partialling-out formula. "orth-IV" is Neyman-orthogonal score with the IV-type formula. Default is "orth-PO".
dml_procedure: (character(1))
A character(1) ("dml1" or "dml2") specifying the double machine learning algorithm. Default is "dml2".
draw_sample_splitting: (logical(1))
Indicates whether the sample splitting should be drawn during initialization of the object. Default is TRUE.
apply_cross_fitting: (logical(1))
Indicates whether cross-fitting should be applied. Default is TRUE.

`xtifedml_plr$set_ml_nuisance_params()`

Set hyperparameters for the nuisance models of DML models with IFE.

Usage

xtifedml_plr$set_ml_nuisance_params(
  learner = NULL,
  treat_var = NULL,
  params,
  set_fold_specific = FALSE
)

Arguments

learner: (character(1))
The nuisance model/learner (see method params_names).
treat_var: (character(1))
The treatment varaible (hyperparameters can be set treatment-variable specific).
params: (named list())
A named list() with estimator parameters. Parameters are used for all folds by default. Alternatively, parameters can be passed in a fold-specific way if option fold_specificis TRUE. In this case, the outer list needs to be of length n_rep and the inner list of length n_folds.
set_fold_specific: (logical(1))
Indicates if the parameters passed in params_theta should be passed in fold-specific way. Default is FALSE. If TRUE, the outer list needs to be of length n_rep and the inner list of length n_folds.

Returns

self

`xtifedml_plr$tune()`

Hyperparameter-tuning within double machine learning.

Usage

xtifedml_plr$tune(
  param_set,
  tune_settings = list(n_folds_tune = 5, rsmp_tune = mlr3::rsmp("cv", folds = 5), measure
    = NULL, terminator = mlr3tuning::trm("evals", n_evals = 20), algorithm =
    mlr3tuning::tnr("grid_search"), resolution = 5),
  tune_on_folds = FALSE
)

Arguments

param_set

(named list())
A named list with a parameter grid for each nuisance model/learner (see method learner_names()). The parameter grid must be an object of class ParamSet.

tune_settings

(named list())
A named list() with arguments passed to the hyperparameter-tuning with mlr3tuning to set up a tuning instance using mlr3tuning::TuningInstanceBatchSingleCrit$new() (see the mlr3tuning package).

tune_settings has entries

terminator (Terminator)
A Terminator object. Specification of terminator is required to perform tuning.
algorithm (Tuner or character(1))
A Tuner object (recommended) or key passed to the respective dictionary to specify the tuning algorithm used in tnr(). algorithm is passed as an argument to tnr(). If algorithm is not specified by the users, default is set to "grid_search". If set to "grid_search", then additional argument "resolution" is required.
rsmp_tune (Resampling or character(1))
A Resampling object (recommended) or option passed to rsmp() to initialize a Resampling for parameter tuning in mlr3. If not specified by the user, default is set to "cv" (cross-validation).
n_folds_tune (integer(1), optional)
If rsmp_tune = "cv", number of folds used for cross-validation. If not specified by the user, default is set to 5.
measure (NULL, named list(), optional)
Named list containing the measures used for parameter tuning. Entries in list must either be Measure objects or keys to be passed to passed to msr(). The names of the entries must match the learner names (see method learner_names()). If set to NULL, default measures are used, i.e., "regr.mse" for continuous outcome variables and "classif.ce" for binary outcomes.
resolution (character(1))
The key passed to the respective dictionary to specify the tuning algorithm used in tnr(). resolution is passed as an argument to tnr().

tune_on_folds

(logical(1))
Indicates whether the tuning should be done fold-specific or globally. Default is FALSE.

Returns

self

`xtifedml_plr$clone()`

The objects of this class are cloneable with this method.

Usage

xtifedml_plr$clone(deep = FALSE)

Arguments

deep: Whether to make a deep clone.

Examples


# An illustrative example using a regression tree (`rpart`)
library(mlr3)
library(rpart)
library(mlr3tuning)

set.seed(1234)

# Generate simulated dataset
data = make_plpr_data(n_obs = 100, t_per = 5, dim_x = 10, theta = 1,
                      r=2, rho=0.6, dgp = "discontinuous")
x_cols  = paste0("X", 1:10)

# Set up DML data environment
obj_xtifedml_data = xtifedml_data_df(data,
                x_cols = x_cols,  y_col = "y", d_cols = "d",
                panel_id = "id",
                time_id = "time",
                approach = "pcce")

# Set up DML estimation environment
 learner = lrn("regr.rpart")
 ml_l = learner$clone()
 ml_m = learner$clone()

 obj_xtifedml = xtifedml_plr$new(obj_xtifedml_data,
                           ml_l = ml_l, ml_m = ml_m,
                           score = "orth-PO", n_folds = 3)
 param_grid = list("ml_l" = ps(cp = p_dbl(lower = 0.01, upper = 0.02),
                            maxdepth = p_int(lower = 2, upper = 10)),
                  "ml_m" = ps(cp = p_dbl(lower = 0.01, upper = 0.02),
                            maxdepth = p_int(lower = 2, upper = 10)))

 tune_settings = list(n_folds_tune = 3,
                   rsmp_tune = mlr3::rsmp("cv", folds = 3),
                   terminator = mlr3tuning::trm("evals", n_evals = 5),
                   tuner = tnr("grid_search", resolution = 10))

 obj_xtifedml$tune(param_set = param_grid, tune_settings = tune_settings)
 obj_xtifedml$fit()

# An illustrative example using a regression tree (`rpart`)
library(mlr3)
library(rpart)
library(mlr3tuning)

set.seed(1234)

# Generate simulated dataset
data = make_plpr_data(n_obs = 100, t_per = 5, dim_x = 10, theta = 1,
                      r=2, rho=0.6, dgp = "discontinuous")
x_cols  = paste0("X", 1:10)

# Set up DML data environment
obj_xtifedml_data = xtifedml_data_df(data,
                x_cols = x_cols,  y_col = "y", d_cols = "d",
                panel_id = "id",
                time_id = "time",
                approach = "pcce")

# Set up DML estimation environment
 learner = lrn("regr.rpart")
 ml_l = learner$clone()
 ml_m = learner$clone()

 obj_xtifedml = xtifedml_plr$new(obj_xtifedml_data,
                           ml_l = ml_l, ml_m = ml_m,
                           score = "orth-PO", n_folds = 3)
 param_grid = list("ml_l" = ps(cp = p_dbl(lower = 0.01, upper = 0.02),
                            maxdepth = p_int(lower = 2, upper = 10)),
                  "ml_m" = ps(cp = p_dbl(lower = 0.01, upper = 0.02),
                            maxdepth = p_int(lower = 2, upper = 10)))

 tune_settings = list(n_folds_tune = 3,
                   rsmp_tune = mlr3::rsmp("cv", folds = 3),
                   terminator = mlr3tuning::trm("evals", n_evals = 5),
                   tuner = tnr("grid_search", resolution = 10))

 obj_xtifedml$tune(param_set = param_grid, tune_settings = tune_settings)
 obj_xtifedml$fit()

Package 'xtifedml'

Help Index

Data generating process for partially linear panel regression models with interactive fixed effects

Description

Usage

Arguments

Value

Examples

Abstract Class xtifedml

Description

Format

Active bindings

Methods

Public methods

xtifedml$new()

Usage

xtifedml$print()

Usage

xtifedml$fit()

Usage

Arguments

Returns

xtifedml$split_samples()

Usage

Returns

xtifedml$tune()

Usage

Arguments

Returns

xtifedml$summary()

Usage

Arguments

xtifedml$confint()

Usage

Arguments

Returns

xtifedml$learner_names()

Usage

Returns

xtifedml$params_names()

Usage

Returns

xtifedml$set_ml_nuisance_params()

Usage

Arguments

Returns

xtifedml$get_params()

Usage

Arguments

Returns

xtifedml$clone()

Usage

Arguments

See Also

Data Setup for Panel Data Approaches with Cluster Variables

Description

Active bindings

Methods

Public methods

xtifedml_data$new()

Usage

Arguments

xtifedml_data$print()

Usage

xtifedml_data$set_data_model()

Usage

Arguments

xtifedml_data$clone()

Usage

Arguments

Wrapper for Double machine learning data-backend initialization from data.frame.

Description

Usage

Arguments

Value

Examples

Routine to estimate partially linear panel regression models with fixed effects within double machine learning.

Description

Format

Details

`xtifedml$new()`

`xtifedml$print()`

`xtifedml$fit()`

`xtifedml$split_samples()`

`xtifedml$tune()`

`xtifedml$summary()`

`xtifedml$confint()`

`xtifedml$learner_names()`

`xtifedml$params_names()`

`xtifedml$set_ml_nuisance_params()`

`xtifedml$get_params()`

`xtifedml$clone()`

`xtifedml_data$new()`

`xtifedml_data$print()`

`xtifedml_data$set_data_model()`

`xtifedml_data$clone()`

`xtifedml_plr$new()`

`xtifedml_plr$set_ml_nuisance_params()`

`xtifedml_plr$tune()`

`xtifedml_plr$clone()`