Skip to content

Factor analysis of multi-neuron spike trains in R

License

Notifications You must be signed in to change notification settings

meixichen/fastr

Repository files navigation

fastr

Meixi Chen, Martin Lysy


Description

The fastr (Factor Analysis of Spike Trains in R) package provides methods for analyzing neural spike trains simultaneously recorded from multiple neurons, with a focus on exploring the correlation between neurons.

Installation

Before installing fastr, make sure you have TMB installed following the instructions here.

To download the development version of this package, run

devtools::install_github("meixichen/fastr")

Factor analysis of multi-neuron spike trains

The main method in this package models the multi-neuron spike trains within a factor analysis framework. Simply put, it assumes that the unobserved neural dynamics \boldsymbol{x}(t)\in\mathbb{R}^q, which determines the observed spike trains \boldsymbol{y}(t)\in\mathbb{R}^q, can be written as \boldsymbol{x}(t)=\boldsymbol{\Lambda}\boldsymbol{f}(t) + \boldsymbol{\epsilon}(t), where \boldsymbol{f}(t)\in\mathbb{R}^d, \ d\ll q contains the independent “factors” and \boldsymbol{\Lambda}_{q\times d} is the factor loading matrix. Model fitting is carried out using the main functionality in this package: fastr_fit().

Usage

More to come here.

Releases

No releases published

Packages

No packages published