Basic Tutorials

• Basic HDDM Tutorial
  • Loading data
  • Fitting a hierarchical model
• Within-subject effects
• Fitting regression models
• Dealing with outliers
• Posterior Predictive Checks
  • Using PPC for model comparison with the groupby argument
  • Defining your own summary statistics
  • Summary statistics relating to outside variables
• Stimulus coding with HDDMRegression
  • Model Recovery Test for HDDMRegression
  • Creating simulated data for the experiment
  • Setting up the HDDM regression model
  • Comparing generative and recovered model parameters
• Fitting go/no-go using the chi-qsuare approach
  • by Jan Willem de Gee (jwdegee@gmail.com)
• LAN Tutorial
  • Things to look out for:
  • Section 0: Colab Prep (Optional)
  • Section 1: Model Info / Simulation / Basic Plotting
  • Section 2: Single Subject (or collapsed) Data
  • Section 3: Hierarchical Models
  • Section 4: Parameter varies by Condition
  • Section 5: Regressors
  • Section 6: Stim Coding
  • Section 7: Model Recovery
  • Section 8: Real Data!
  • Section 9: Accessing the Neural Network Directly
• Tutorial on Parameter defaults
  • Install (colab)
  • Load Modules
  • Example Models
• Tutorial on the “Model Plot”
  • Install (colab)
  • Import modules
  • Example 1: Simple Data from ddm_hddm_base model
  • Example 2: Simple Data from a model enabled by LAN-extension
  • Example 3: More complex data
  • Example 4: Making gifs
• Tutorial for analyzing instrumental learning data with the HDDMrl module
  • OUTLINE
  • 1. Background
  • 2. Installing the module
  • 3. How the RLDDM works
  • 4. Structuring data
  • 5. Running basic model
  • 6. Checking results
  • 7. Posterior predictive checks
  • 8. Parameter recovery
  • 9. Separate learning rates for positive and negative prediction errors
  • Posterior predictive check
  • 10. depends_on vs. split_by
  • 11. Probabilistic binary outcomes vs. normally distributed outcomes
  • 12. HDDMrlRegressor
  • 13. Regular RL without RT
  • Posterior predictive check
• Tutorial for analyzing instrumental learning data with the HDDMnnRL module
  • Load the data
  • Initialize the HDDMnnRL model and sample
  • Save the model
  • Check the posterior results
  • Posterior Predictive Checks