Data and analysis code for investigation into neural activity in auditory cortex of ferrets during perceptual constancy. The best way to use this repository is to read the paper (available here) and then look at the scripts for details of specific analyses.

The majority of data analysis was performed in Matlab (version 2014 or later) but the data was collected using GoFerret 1.0 and Matlab 2013b. For advanced analysis with original data held in TDT file types, you will also need to install OpenDeveloper (Tucker Davis Technologies).

For running live scripts, you'll need to add the 'lib' directory to the Matlab path to use helper functions, e.g.:

addpath('home/stephen/GitHub/perceptual_constancy_for_vowels/lib')

• Plot sound spectra (notebook)
• Plot behavioral performance (notebook)
• Visualize activity of individual neurons

• Start with the ReadMe that introduces the motivation and design of decoders for the current study.
• Decode sounds presented to listeners from individual neuron activity (Script)
• Predict future choices of listeners from individual neuron activity (Script)
• Decode sounds presented to listeners from activity of neural populations (Script)

• Compare hyperparameters that best decode different sound features (Script)
• Compare decoding of neural activity in task engaged and passively listening animals (Script)
• Compare decoding of neural activity in animals with and without behavioral training (Script)

For those users who are interested in replicating or extending the study by collecting new data:

• Generate stimuli
• Run behavioral task using GoFerret
• Run spike extraction pipeline