Simulation and Modeling of Variable Precision Model

• The Modelling the Working memory resource allocation

This repository is a fork of the original repository from Yoo Aspen (https://github.com/aspenyoo/WM_resource_allocation). The original parent repository is accompanied from Yoo AH's paper paper Strategic allocation of working memory resource by Yoo, Klyszejko, Curtis, & Ma (2018, Sci Reports). [link]

I created this repository as the result of my simulation and modeling of the variable precision model based on Yoo's codes. I also write a documentation of a introduction of variable precision model, and my simulation and modeling practices as a review of the VP model, which i could share with u if u contact me by qy775@nyu.edu

---

My self contributions are:

• simulate&model_QY.m: the codes for simulation and modeling practice
  • Simulate data from evenly allocation VP model parameters, with diff params combinations.
  • Simulate from proportional allocation VP model parameters, with fixed Jbar or fixed tau.
  • Fit real data with negative Log Likelihood calculation.
  • Recover the parameter combination (theta2) from the simulated data (theta1)
• Proportional_VP_single_simulator_QY: the function for simulate a error data by Proportional VP model with a params combination [Jbar, tau]
• Proportional_calc_nll_QY.m: （revision needed） the function for calculate the neg log-likelihood based on a data and Propor_VP model with a params combination [Jbar, tau], works for 1 by 3 cells while not 1 by 1 cell.
• Proportional_fitparams_QY.m: （revision needed） the function for get the best fitting param combination from the data, by certain runs of calculations, works for 1 by 3 cells while not 1 by 1 cell.
• qy_modelling results: contains results of my model simulation, parameter recovery, model comparison

---

A brief description of the orginal repository organization (details are found within each file):

• data/priority: contains data for the first and second experiment.
• fits/priority: contains model predictions for the first and second experiment.
• helperfunctions: functions necessary for model fitting, plots, etc.
• model: all functions necessary to fit any of the models to data.
• tutorial: this contains files for a lab meeting tutorial. It may not be helpful for you.
• plots_for_pub.m: recreate all plots in publication
• stats_for_pub.m redo stats reported in publication