Function to estimate autocorrelation parameters using dichotomized Gaussian simulations

This function performs the same estimate-and-fit procedure as process.autocorr, but does so multiple times for each neuron using simulated spike rasters generated from a dichotomized Gaussian model of that neuron.

Usage

estimate.autocorr.params(
  neuron_list,
  n_trials_per_sim = 300,
  n_sims_per_neurons = 100,
  bin_count_action = "sum",
  max_lag = 0,
  A0 = 0.001,
  tau0 = 1,
  ctol = 1e-08,
  max_evals = 500,
  use_raw = TRUE
)

Arguments

neuron_list: An R list of neuron objects.
n_trials_per_sim: Number of trials to simulate for each simulation (default: 300).
n_sims_per_neurons: Number of simulations to run for each neuron (default: 100).
bin_count_action: Method for counting spikes in each bin when computing autocorrelation; one of "boolean", "mean", or "sum" (default: "sum").
A0: Initial guess for amplitude parameter of exponential decay function (default: 0.001).
tau0: Initial guess for time constant parameter of exponential decay function (default: 1.0).
ctol: Convergence tolerance for fitting exponential decay function (default: 1e-8).
max_evals: Maximum number of evaluations for fitting exponential decay function (default: 500).
use_raw: Logical indicating whether to use raw autocorrelation (true) or standard centered and normalized correlation (false) (default: TRUE).

Value

A list containing a data frame of autocorrelation parameter estimates (one row per simulation), a data frame of neuron identifiers (one row per neuron), and the number of simulations run per neuron.