Introduction to bkmr and bkmrhat | How to use the bkmrhat package | Example 1: single vs multi-chains | Example 2: Diagnostics | Example 3: Posterior summaries | Example 4: diagnostics and inference when variable selection is used (Bayesian model averaging over the scale parameters of the kernel function) | Example 5: Parallel posterior summaries as diagnostics | Example 6: Continuing a fit | Acknowledgments

Introduction

Table of Contents | Example 7: time-to-event analysis and parallel processing | Example 8: clustering | Example 9: partial effects | Example 10: multinomial outcomes | Example 11: sample weighting from, e.g. NHANES | Missing data, limits of detection and multiple imputation | Limits of detection | Multiple imputation | References | Acknowledgments

Table of Contents | Introduction | The model | How to use the qgcomp package | Example 1: linear model | Example 2: conditional odds ratio, marginal odds ratio in a logistic model | Example 3: adjusting for covariates, plotting estimates | Example 4: non-linearity (and non-homogeneity) | Aside: some details on qgcomp methods for non-linearity | Interpretation of model parameters | Example 5: comparing model fits and further exploring non-linearity | Graphical approach to explore non-linearity in a correlated subset of exposures using splines | Caution about graphical approaches | Example 6: miscellaneous other ways to allow non-linearity | using indicator terms for each quantile | interactions between indicator terms | breaks at specific quantiles (these breaks act on the quantized basis) | Statistical approach explore non-linearity in a correlated subset of exposures using splines | FAQ | Why don't I get weights/scaled effects from the boot or ee functions? (and other questions about the weights/scaled effect sizes) | Do I need to model non-linearity and non-additivity of exposures? | Do I have to use quantiles? | Can I cite this document? | Where else can I get help? | References | Acknowledgments

Table of Contents | Introduction | The model | Basics: fitting a model with a modifier | Basics: getting bounds for pointwise comparisons | Basics: plotting weights (weights are at referent level of modifier) | Basics: bootstrapping | Plotting bootstrapped fits (plotting predictions) | Overlay plots of predictions at multiple modifier levels | Basics: Estimating equations | Plotting predictions from estimating equation fits (plotting predictions) | Basics: categorical modifier, binary outcome | Simulated data defaults | The wrong way to include categorical modifiers | The right way to include categorical modifiers (use as.factor()) | Bounds for pointwise comparisons with categorical modifiers | Overlay plots of predictions at multiple modifier levels | Global tests of interaction for categorical modifiers | Basics: Non-numerical factor modifiers | Basics: Continuous modifiers | Stratified effects, weights, and pointwise effect comparisons at specific values of a continuous confounder | Non-linear fits | Survival analysis | Frequently asked questions | How does qgcomp/qgcompint address collinearity | Is there an upper limit in terms of the number of variables I can include? | Why are estimating equation methods and bootstrapping both used? Is one preferable? | Why isn't available in this package even though it is in the qgcomp package? | How do I use multiple modifiers? | See also | References | Original quantile g-computation paper | First paper to use qgcompint | Acknowledgments

A guided walk through the Metropolis algorithmm | The data | Helper functions | Maximum likelihood estimates | Random walk metropolis | Inspecting output | Guided metropolis | Contrasting output with random walk | Guided, adaptive metropolis algorithm | Contrasting output | Guided, adaptive metropolis algorithm using normal priors | Using the R package | Using the R package, smart initial values | Extending the logistic model results after samples are generated