Getting Started with RMeDPower2 • RMeDPower2

Introduction

RMeDPower2 is an R package for statistical power analysis in repeated measures experiments. It provides complete functionality to analyze data from hierarchical or crossed experimental designs, test modeling assumptions, identify outliers, estimate statistical power, and perform sample size calculations.

Installation

You can install RMeDPower2 from GitHub:

# Install devtools if you haven't already
install.packages("devtools")

# Install RMeDPower2 from GitHub
devtools::install_github("gladstone-institutes/RMeDPower2", build_vignettes = TRUE)

Quick Start

Here’s a basic example of using RMeDPower2:

library(RMeDPower2)

# Load example data
data(RMeDPower_data1)

# Define experimental design
design <- new("RMeDesign",
    response_column = "cell_size2",
    condition_column = "classification",
    experimental_columns = c("experiment", "line"),
    condition_is_categorical = TRUE
)

# Define probability model
model <- new("ProbabilityModel",
    error_is_non_normal = FALSE
)

# Define power parameters
power_param <- new("PowerParams",
    target_columns = c("experiment"),
    power_curve = 1,
    nsimn = 100
)

# Step 1: Diagnose data and model assumptions
diagnose_res <- diagnoseDataModel(RMeDPower_data1, design, model)

# Step 2: Calculate power
power_res <- calculatePower(RMeDPower_data1, design, model, power_param)

# Step 3: Get parameter estimates
estimate_res <- getEstimatesOfInterest(RMeDPower_data1, design, model)

Key Features

Experimental Design Support: Handles nested and crossed experimental designs
Power Analysis: Simulation-based power calculations using mixed-effects modeling
Outlier Detection: Two-level outlier identification at observation and group levels
Model Validation: Comprehensive assumption testing and diagnostic visualizations
Flexible Distributions: Support for normal, binomial, Poisson, and negative binomial distributions

Next Steps

Read the Tutorial for a complete walkthrough
See Examples for real-world applications
Check the Reference for detailed function documentation