Welcome to getting structural equation modeling done.
umx is an R package to help you get things done with Structural Equation Modeling using OpenMx. The design goals are to make even complex SEM - like twin models, of genomicSEM with WLS quicker and easier to implement, update, and publish.
Chapters
- Collected introductory umx articles
- Gallery of Example Models!
- Collected in-depth RAM function articles
- Collected articles on twin modeling in umx
- Collected technical articles (parallel, speed, matrix-algebra)
- Installing umx (1-minute read)
- Build & run your first SEM model
- What you expect is what you get
- Taking Stock: What you’ve seen so far
- umxPath
- plotting
- umxSummary
- Values and why they are important
- Labels to fix, drop, or equate paths.
- umxModify & umxCompare
- Factor scores with missing data!
- Multiple groups
- Definition variables in umxRAM
- Get more confidence (intervals), easily!
- Constraints
- umxRAM with Ordinal Data
- Confidence on standardized parameters
- Statistical helper functions
- Cholesky paths with umxPath
- Twin ACE Cholesky with umxACE
- EFA (Exploratory factor analysis) using umx
- Mendelian Randomisation and Instrumental Variables in umx
- Growth Curve
- Power
- multi-level modeling in OpenMx
- Simulation
- Reflective vs Formative
- Data Wrangling (scale, rename, residualize data, twin data, umxAPA)
- Linear Algebra
- Parallel Execution
- Set the Optimizer using umx_set_optimizer
- Matrix concepts: Not positive definite, Hessian, Jacobian