Group-Based Finite Mixture Models: A Latent Trajectory Approach

Code for course project use only at this time.

Final project for STA 663

• Takes time variable and continuous outcome
• Generates latent trajectories (time-dependent clusters) of outcome behavior over time

Dependencies

Required

• Python (created using 2.7.x)
• Numpy
• Scipy
• Pymix (https://github.com/klahnakoski/pymix.git)

Recommended

• ggplot (for trajectory plotting function included in the repository)
• pandas (for trajectory plotting function included in the repository)
• STAN (for Bayesian alternative specification)
• pystan (for Bayesian alternative specification)