Learn R Programming

⚠️There's a newer version (1.4.1) of this package.Take me there.

Reference Based Multiple Imputation (`rbmi`)

Overview

The rbmi package is used for the imputation of missing data in clinical trials with continuous multivariate normal longitudinal outcomes. It supports imputation under a missing at random (MAR) assumption, reference-based imputation methods, and delta adjustments (as required for sensitivity analysis such as tipping point analyses). The package implements both Bayesian and approximate Bayesian multiple imputation combined with Rubin's rules for inference, and frequentist conditional mean imputation combined with (jackknife or bootstrap) resampling.

Installation

The package can be installed directly from CRAN via:

install.packages("rbmi")

Note that the usage of Bayesian multiple imputation requires the installation of the suggested package rstan.

install.packages("rstan")

Usage

The package is designed around its 4 core functions:

draws() - Fits multiple imputation models
impute() - Imputes multiple datasets
analyse() - Analyses multiple datasets
pool() - Pools multiple results into a single statistic

The basic usage of these core functions is described in the quickstart vignette:

vignette(topic = "quickstart", package = "rbmi")

Validation

For clarification on the current validation status of rbmi please see the FAQ vignette.

Support

For any help with regards to using the package or if you find a bug please create a GitHub issue

Copy Link

Version

Install

install.packages('rbmi')

Monthly Downloads

368

Version

1.3.1

License

Apache License (>= 2)

Issues

Pull Requests

Stars

Forks

Repository

https://github.com/insightsengineering/rbmi

Homepage

https://insightsengineering.github.io/rbmi/

Maintainer

Craig Gower-Page

Last Published

December 11th, 2024

Functions in rbmi (1.3.1)

Applies delta adjustment

QR decomposition

adjust_trajectories_single

Adjust trajectory of a subject's outcome due to the intercurrent event (ICE)

Analysis of Covariance

Implements an Analysis of Covariance (ANCOVA)

R6 Class for a FIFO stack

as_cropped_char

as_cropped_char

adjust_trajectories

Adjust trajectories due to the intercurrent event (ICE)

Construct an analysis object

Convert object to dataframe

assert_variables_exist

Assert that all variables exist within a dataset

Convert character variables to factor

Creates a draws object

Expand and fill in missing data.frame rows

Analyse Multiple Imputed Datasets

Creates a "MMRM" ready dataset

Create an imputation object

extract_covariates

Extract Variables from string vector

Expand data.frame into a design matrix

as_mmrm_formula

Create MMRM formula

as_simple_formula

Creates a simple formula object from a string

Compute covariance matrix for some reference-based methods (JR, CIR)

Diagnostics of the MCMC

Convert indicator to index

Create a delta data.frame template

Create vector of Stratas

check_hmc_diagn

Diagnostics of the MCMC based on HMC-related measures.

Diagnostics of the MCMC based on ESS

Ensure rstan exists

Fit the base imputation model and get parameter estimates

convert_to_imputation_list_df

Convert list of imputation_list_single() objects to an imputation_list_df() object (i.e. a list of imputation_df() objects's)

Evaluate a call to mmrm

extract_imputed_df

Extract imputed dataset

Get imputation strategies

Calculate delta from a lagged scale coefficient

extract_data_nmar_as_na

Set to NA outcome values that would be MNAR if they were missing (i.e. which occur after an ICE handled using a reference-based imputation strategy)

Extract draws from a stanfit object

extract_imputed_dfs

Extract imputed datasets

get_bootstrap_stack

Creates a stack object populated with bootstrapped samples

Extract the Effective Sample Size (ESS) from a stanfit object

get_conditional_parameters

Derive conditional multivariate normal parameters

Fit a MMRM model

get_pattern_groups

Determine patients missingness group

get_pool_components

Expected Pool Components

Extract parameters from a MMRM model

get_pattern_groups_unique

Get Pattern Summary

generate_data_single

Generate data for a single group

Get Compiled Stan Object

get_visit_distribution_parameters

Derive visit distribution parameters

Fit the base imputation model using a Bayesian approach

get_jackknife_stack

Creates a stack object populated with jackknife samples

Is single character

Von Hippel and Bartlett pooling of BMLMI method

get_session_hash

Get session hash

get_mmrm_sample

Fit MMRM and returns parameter estimates

get_example_data

Simulate a realistic example dataset

Does object have a class ?

get_delta_template

Get delta utility variables

is_in_rbmi_development

Is package in development mode?

imputation_list_single

A collection of imputation_singles() grouped by a single subjid ID

is_num_char_fact

Is character, factor or numeric

Last Observation Carried Forward

Create imputed datasets

Fit the base imputation model on bootstrap samples

imputation_single

Create a valid imputation_single object

impute_internal

Create imputed datasets

Calculate parametric confidence intervals

impute_data_individual

Impute data for a single subject

Sample outcome value

Create a valid imputation_df object

longDataConstructor

R6 Class for Storing / Accessing & Sampling Longitudinal Data

imputation_list_df

List of imputations_df

Is value absent

Set the multiple imputation methodology

Pool analysis results obtained from the imputed datasets

Calculate design vector for the lsmeans

print.imputation

Print imputation object

Is character or factor

Parallelise Lapply

pval_percentile

P-value of percentile bootstrap

pool_bootstrap_normal

Bootstrap Pooling via normal approximation

random_effects_expr

Construct random effects formula

Invert and derive indexes

pool_bootstrap_percentile

Bootstrap Pooling via Percentiles

R6 Class for printing current sampling progress

Create object of sample_single class

Least Square Means

scalerConstructor

R6 Class for scaling (and un-scaling) design matrices

make_rbmi_cluster

Create a rbmi ready cluster

remove_if_all_missing

Remove subjects from dataset if they have no observed values

Barnard and Rubin degrees of freedom adjustment

Internal Pool Methods

rbmi: Reference Based Multiple Imputation

prepare_stan_data

Prepare input data to run the Stan model

Print analysis object

Capture all Output

Set simulation parameters of a study group.

recursive_reduce

recursive_reduce

split_imputations

Split a flat list of imputation_single() into multiple imputation_df()'s by ID

Print draws object

Set key variables

Combine estimates using Rubin's rules

validate.is_mar

Validate is_mar for a given subject

Does a string contain a substring

transpose_imputations

Transpose imputations

transpose_results

Transpose results object

Sample Patient Ids

Create and validate a sample_list object

validate_datalong

Validate a longdata object

Validate inputs for vars

validate_strategies

Validate user specified strategies

Sample random values from the multivariate normal distribution

Simulate intercurrent event

simulate_test_data

Create simulated datasets

simulate_dropout

Simulate drop-out

Sort data.frame

validate.analysis

Validate analysis objects

Transform array into list of arrays

transpose_samples

Transpose samples

Validate draws object

Generic validation method

validate.stan_data

Validate a stan_data object

validate.references

Validate user supplied references

validate.sample_list

Validate sample_list object

validate.sample_single

Validate sample_single object

validate.simul_pars

Validate a simul_pars object

validate_analyse_pars

Validate analysis results

antidepressant_data

Antidepressant trial data