Model Audit - Verification, Validation, and Error Analysis

Provides an easy to use unified interface for creating validation plots for any model. The 'auditor' helps to avoid repetitive work consisting of writing code needed to create residual plots. This visualizations allow to asses and compare the goodness of fit, performance, and similarity of models.

Error Grid Analysis

Functions for assigning Clarke or Parkes (Consensus) error grid zones to blood glucose values, and for plotting both types of error grids in both mg/mL and mmol/L units.

Tools for Environmental and Ecological Survey Design

Statistical tools for environmental and ecological surveys. Simulation-based power and precision analysis; detection probabilities from different survey designs; visual fast count estimation.

Bounding Treatment Effects by Limited Information Pooling

Estimation and inference methods for bounding average treatment effects (on the treated) that are valid under an unconfoundedness assumption. The bounds are designed to be robust in challenging situations, for example, when the conditioning variables take on a large number of different values in the observed sample, or when the overlap condition is violated. This robustness is achieved by only using limited "pooling" of information across observations. For more details, see the paper by Lee and Weidner (2021), "Bounding Treatment Effects by Pooling Limited Information across Observations," .

Single-Index Models with a Surface-Link

An implementation of a single-index regression for optimizing individualized dose rules from an observational study. To model interaction effects between baseline covariates and a treatment variable defined on a continuum, we employ two-dimensional penalized spline regression on an index-treatment domain, where the index is defined as a linear combination of the covariates (a single-index). An unspecified main effect for the covariates is allowed, which can also be modeled through a parametric model. A unique contribution of this work is in the parsimonious single-index parametrization specifically defined for the interaction effect term. We refer to Park, Petkova, Tarpey, and Ogden (2020) (for the case of a discrete treatment) and Park, Petkova, Tarpey, and Ogden (2021) "A single-index model with a surface-link for optimizing individualized dose rules" for detail of the method. The model can take a member of the exponential family as a response variable and can also take an ordinal categorical response. The main function of this package is simsl().

Bayesian Model Selection in Logistic Regression for the Detection of Adverse Drug Reactions

Spontaneous adverse event reports have a high potential for detecting adverse drug reactions. However, due to their dimension, the analysis of such databases requires statistical methods. We propose to use a logistic regression whose sparsity is viewed as a model selection challenge. Since the model space is huge, a Metropolis-Hastings algorithm carries out the model selection by maximizing the BIC criterion.

Correct Bias in Estimated Regression Coefficients

This function corrects the bias in estimated regression coefficients due to classical additive measurement error (i.e., within-person variation) in logistic regressions under the main study/external reliability study design and the main study/internal reliability study design. The output includes the naive and corrected estimators for the regression coefficients; for the variance estimates of the corrected estimators, the extra variation due to estimating the parameters in the measurement error model is ignored or taken into account. Reference: Carroll RJ, Ruppert D, Stefanski L, Crainiceanu CM (2006) .

Adaptive Semiparametric Additive Regression with Simultaneous Confidence Bands and Specification Tests

Fits semiparametric additive regression models with spatially adaptive penalized splines and computes simultaneous confidence bands and associated specification (lack-of-fit) tests. Simultaneous confidence bands cover the entire curve with a prescribed level of confidence and allow us to assess the estimation uncertainty for the whole curve. In contrast to pointwise confidence bands, they permit statements about the statistical significance of certain features (e.g. bumps) in the underlying curve.The method allows for handling of spatially heterogeneous functions and their derivatives as well as heteroscedasticity in the data. See Wiesenfarth et al. (2012) .

Smooth-Rough Partitioning of the Regression Coefficients

Performs the change-point detection in regression coefficients of linear model by partitioning the regression coefficients into two classes of smoothness. The change-point and the regression coefficients are jointly estimated.

Methods for Transition Probabilities

Provides a function for estimating the transition probabilities in an illness-death model. The transition probabilities can be estimated from the unsmoothed landmark estimators developed by de Una-Alvarez and Meira-Machado (2015) . Presmoothed estimates can also be obtained through the use of a parametric family of binary regression curves, such as logit, probit or cauchit. The additive logistic regression model and nonparametric regression are also alternatives which have been implemented. The idea behind the presmoothed landmark estimators is to use the presmoothing techniques developed by Cao et al. (2005) in the landmark estimation of the transition probabilities.