aIc and amIcomp implement simple test functions to determine if a particular dataset and transformation exhibit data pathologies that we can understand and control for. Since 'all models are wrong'[Box:1976] it is important to know at the outset what is wrong, why it is wrong, and how that wrongness would hinder general interpretability in some way. | List of common data pathologies: | Properties of datasets | Example functions | Sub-compositions and correlation (coherence) | Sub-compositions and distance (dominance) | Testing for scale invariance | Testing for perturbation invariance | Choosing parameters | Conclusion | SessionInfo

Introduction to r Biocpkg("ALDEx2") | Installation | What r Biocpkg("ALDEx2") does differently | Quick Start: aldex with 2 groups: | Modular r Biocpkg("ALDEx2") is way more informative | The aldex.clr module | The aldex.ttest module | The aldex.effect module | The aldex.plot module | The effect confidence interval | Complex study designs and the aldex.glm module | The aldex.kw module | Adding scale to ALDEx2 | ALDEx2 outputs | Expected values | Explaining the outputs | aldex.effect(x) | aldex.ttest(x) | aldex.glm(x) | A word about effect size and overlap | Alternative plotting of outputs | Troubleshooting datasets | Using scale to correct for asymmetric datasets | Subsetting features to correct for asymmetric datasets | Parameter values for denom | Very large datasets | Contributors | Version information

Introduction to Scale Simulation using ALDEx2 | From Normalizations to Scale Models | Using Scale Simulation in ALDEx2 | Installing ALDEx2 with Scale Simulation | Simulation Setup | Incorporating Scale in ALDEx2 | Default Scale Model | Measurement-Based Scale Model | Knowledge-Based Scale Model | Comparing the Models | Sensitivity Analyses | Session Info | References