Title: Robust Inference in Complex Innovative Trial Design
Version: 1.0.0
Description: Perform robust estimation and inference in platform trials and other master protocol trials. Yuhan Qian, Yifan Yi, Jun Shao, Yanyao Yi, Gregory Levin, Nicole Mayer-Hamblett, Patrick J. Heagerty, Ting Ye (2025) <doi:10.48550/arXiv.2411.12944>.
License: MIT + file LICENSE
Encoding: UTF-8
RoxygenNote: 7.3.2
URL: https://github.com/Eureeca/RobinCID
BugReports: https://github.com/Eureeca/RobinCID/issues
Suggests: testthat (≥ 3.0.0)
Config/testthat/edition: 3
Imports: checkmate, numDeriv, stats, MASS
Depends: R (≥ 3.5)
LazyData: true
NeedsCompilation: no
Packaged: 2025-05-27 06:40:54 UTC; eureka
Author: Yuhan Qian [aut, cre, cph], Ting Ye [aut], Yanyao Yi [aut], Marlena Bannick [aut]
Maintainer: Yuhan Qian <yhqian@uw.edu>
Repository: CRAN
Date/Publication: 2025-05-29 08:50:13 UTC

RobinCID Package

Description

RobinCID implements unbiased prediction and robust inference in R.

Author(s)

Maintainer: Yuhan Qian yhqian@uw.edu [copyright holder]

Authors:

See Also

Useful links:

Assign Probability according to Design

Description

Assign Probability according to Design

Usage

assign_prob_and_strata(
  data,
  estimand,
  design = list(randomization_var_colnames = NULL, randomization_table = NULL),
  method,
  estimated_propensity = TRUE,
  stratify_by = NULL
)

Arguments

data

(data.frame) Input data frame.

estimand

(list) A list specifying the estimand.

design

(list) A list describing the randomization design. See Details.

method

estimation method.

estimated_propensity

Whether to use estimated propensity score.

stratify_by

The column name of stratification variable in data.

Details

design has two elements: randomization_var_colnames (vector) and randomization_table (data.frame)

Value

A new data with columns of the treatment assignment probability.

Compute Estimates and Covariance Matrix

Description

Compute Estimates and Covariance Matrix

Usage

estimate_effect(
  ret,
  y,
  treatment,
  treatments_for_compare,
  data,
  prob_mat,
  post_strata,
  stabilize
)

Arguments

ret

counterfactual prediction

y

Observed outcome

treatment

name of treatment

treatments_for_compare

description

data

(data.frame) data

prob_mat

(data.frame) treatment assignment probability

post_strata

(character) A string name of post-stratification variable

stabilize

(logical) whether to stabilize

Value

A list of "estimate_effect" object with following elements:

Trial Data

Description

This dataset is generated based on the first three enrollment windows outlined in Figure 1(a) in our paper.

Usage

example

Format

A data frame with 500 rows and 17 columns:

xc, xb, subtype

The covariates.

t, substudy

The enrollment window and substudy assignment

treatment

The treatment assignment, "1", "2", "3" and "4"

y

The continuous response.

y_b

The binary response, I(y>3).

s12, s12.2, s12.error, s13, s14

The stratification variables.

trt.1, trt.2, trt.3, trt.4

The assgiment probabilities of trt 1, 2, 3, and 4.

Source

The data is generated by Yuhan.

Find Data in a Fit

Description

Find Data in a Fit

Usage

find_data(fit, ...)

Arguments

fit

A fit object.

...

Additional arguments.

Value

A data frame used in the fit.

Contrast Functions and Jacobians

Description

Contrast Functions and Jacobians

Usage

h_diff(x)

h_jac_diff(x)

h_ratio(x)

h_jac_ratio(x)

h_odds_ratio(x)

h_jac_odds_ratio(x)

Arguments

x

(numeric) Vector of values.

Value

Vector of contrasts, or matrix of jacobians.

Examples

h_diff(1:3)
h_jac_ratio(1:3)

Lower Triangular Index

Description

Lower Triangular Index

Usage

h_lower_tri_idx(n)

Arguments

n

(int) Number of rows/columns.

Value

Matrix of lower triangular indices.

Counterfactual Prediction

Description

Obtain counterfactual prediction of a fit.

Usage

predict_counterfactual(
  fit.j,
  fit.k,
  treatment,
  treatments_for_compare,
  prob_mat,
  post_strata,
  data,
  stabilize,
  settings
)

Arguments

fit.j

fitted object for trt j.

fit.k

fitted object for trt k.

treatment

name of treatment column

treatments_for_compare

(character) Treatments for comparison

prob_mat

(data.frame) treatment assignment probabilities

post_strata

(character) A string name of post-stratification variable.

data

(data.frame) raw dataset.

stabilize

stabilize

settings

estimation setting

Value

A list of prediction_cf object with following elements:

S3 Methods for prediction_cf

Description

S3 Methods for prediction_cf

Usage

## S3 method for class 'prediction_cf'
print(x, ...)

Arguments

x

(prediction_cf)
the obtained counter-factual prediction object.

Value

No return value, called for side effects

Functions

Post-Stratification Based Inference

Description

Provides robust inference via post stratification.

Usage

robin_ps(
  data,
  estimand = list(tx_colname = NULL, tx_to_compare = NULL),
  design = list(randomization_var_colnames = NULL, randomization_table = NULL),
  stratify_by = NULL,
  outcome_model = list(formula = NULL, family = gaussian()),
  contrast_specs = list(contrast = "difference", contrast_jac = NULL),
  alpha = 0.05,
  ...
)

Arguments

data

(data.frame) A data frame containing the dataset.

estimand

(list) A list specifying the estimand, with two elements:

  • tx_colname (character): The column name of the treatment variable in data.

  • tx_to_compare (⁠character vector⁠): A vector specifying exactly two treatment levels to compare.

design

(list) A list specifying randomization information, with two elements:

  • randomization_var_colnames (⁠character vector⁠): Column names of randomization variables in data.

  • randomization_table (data.frame, default: NULL): A data frame containing treatment assignment probabilities for each level of the randomization variables. See Details.

stratify_by

(character, optional) The column name of the stratification variable in data. If provided, stratify_by overrides design.

outcome_model

(list) A list specifying the outcome working model, with two elements:

  • formula (formula): The regression formula for the analysis.

  • family A description of the error distribution and link function for the model. Default: gaussian().

contrast_specs

(list) A list specifying the contrast function and its Jacobian:

  • contrast (function or character): A function to compute the treatment effect, or one of "difference", "risk_ratio", or "odds_ratio" for default contrasts.

  • contrast_jac (function, optional): A function to compute the Jacobian of the contrast function. Ignored if using default contrasts.

alpha

(numeric) The nominal significance level. Default: 0.05.

...

Additional arguments passed to glm.

Details

If family is MASS::negative.binomial(NA), the function will use MASS::glm.nb instead of glm.

Value

A treatment_effect object.

Examples

data_sim <- RobinCID::example
tx_colname <- "treatment"
treatment_levels <- unique(data_sim[[tx_colname]])
tx_to_compare <- c("trt.1", "trt.3")
randomization_var_colnames <- c("t", "subtype")
df <- data_sim[c("xb", "xc", tx_colname, randomization_var_colnames, "y")]
randomization_table <- unique(data_sim[c(randomization_var_colnames, treatment_levels)])
robin_ps(
  data = df,
  estimand = list(tx_colname = tx_colname,
                  tx_to_compare = tx_to_compare),
  design = list(randomization_var_colnames = randomization_var_colnames,
                randomization_table = randomization_table),
  stratify_by = NULL,
  outcome_model = list(formula = y ~ 1,
                       family = gaussian())
)

Inverse Probability Weighting Based Inference

Description

Provides robust inference via inverse probability weighting.

Usage

robin_wt(
  data,
  estimand = list(tx_colname = NULL, tx_to_compare = NULL),
  design = list(randomization_var_colnames = NULL, randomization_table = NULL),
  estimated_propensity = TRUE,
  outcome_model = list(formula = NULL, family = gaussian()),
  contrast_specs = list(contrast = "difference", contrast_jac = NULL),
  alpha = 0.05,
  ...
)

Arguments

data

(data.frame) A data frame containing the dataset.

estimand

(list) A list specifying the estimand, with two elements:

  • tx_colname (character): The column name of the treatment variable in data.

  • tx_to_compare (⁠character vector⁠): A vector specifying exactly two treatment levels to compare.

design

(list) A list specifying randomization information, with two elements:

  • randomization_var_colnames (⁠character vector⁠): Column names of randomization variables in data.

  • randomization_table (data.frame, default: NULL): A data frame containing treatment assignment probabilities for each level of the randomization variables. See Details.

estimated_propensity

(logical, default: TRUE) Whether to use estimated propensity scores.

outcome_model

(list) A list specifying the outcome working model, with two elements:

  • formula (formula): The regression formula for the analysis.

  • family A description of the error distribution and link function for the model. Default: gaussian().

contrast_specs

(list) A list specifying the contrast function and its Jacobian:

  • contrast (function or character): A function to compute the treatment effect, or one of "difference", "risk_ratio", or "odds_ratio" for default contrasts.

  • contrast_jac (function, optional): A function to compute the Jacobian of the contrast function. Ignored if using default contrasts.

alpha

(numeric) The nominal significance level. Default: 0.05.

...

Additional arguments passed to glm.

Details

If randomization_table is provided, it must include columns corresponding to randomization_var_colnames, as well as treatment assignment probability columns named after the treatment levels in tx_colname from data.

If family is MASS::negative.binomial(NA), the function will use MASS::glm.nb instead of glm.

Value

A treatment_effect object.

Examples

data_sim <- RobinCID::example
tx_colname <- "treatment"
treatment_levels <- unique(data_sim[[tx_colname]])
tx_to_compare <- c("trt.1", "trt.3")
randomization_var_colnames <- c("t", "subtype")
df <- data_sim[c("xb", "xc", tx_colname, randomization_var_colnames, "y")]
randomization_table <- unique(data_sim[c(randomization_var_colnames, treatment_levels)])
robin_wt(
  data = df,
  estimand = list(tx_colname = tx_colname,
                  tx_to_compare = tx_to_compare),
  design = list(randomization_var_colnames = randomization_var_colnames,
                randomization_table = randomization_table),
  estimated_propensity = FALSE,
  outcome_model = list(formula = y ~ 1,
                       family = gaussian())
)

Treatment Effect

Description

Obtain treatment effect and variance from counter-factual prediction

Usage

treatment_effect(object, pair, eff_measure, eff_jacobian, alpha, ...)

difference(object, ...)

risk_ratio(object, ...)

odds_ratio(object, ...)

Arguments

object

Object from which to obtain treatment effect. Must be obtained from estimate_effect().

pair

(integer or character) Names or index of the treatment levels.

eff_measure

(function) Treatment effect measurement function.

eff_jacobian

(function) Treatment effect jacobian function.

alpha

Nominal level

...

Additional arguments passed to glm

Value

A list of treatment_effect object with following elements: