ZilanyBruceCarney2014.jl

ZilanyBruceCarney2014.jl is a Julia package that provides access to the Zilany, Bruce, and Carney (2014) auditory-nerve model in Julia.

Zilany, M. S. A., Bruce, I. C., & Carney, L. H. (2014). Updated parameters and expanded simulation options for a model of the auditory periphery. The Journal of the Acoustical Society of America, 135(1), 283–286. http://dx.doi.org/10.1121/1.4837815

Why a new package? Why Julia?

Existing bindings for some models already exist in several MATLAB or Python packages such as the original code release, cochlea and Auditory Modeling Toolbox. This package was written in Julia to supplement existing packages and leverage several unique benefits of Julia:

• Interoperability between Julia and C is excellent. Julia has a "no boilerplate" philosophy that results in C bindings that easy to write and maintain.
• Julia can be exceptionally fast, but it is still a high-level langauge. This allows us to extend tools and models written in low-level langauges (such as C) using a high-level language (Julia) while avoiding some of the performance penalty usually associated with writing in a high-level language.
• Julia has a rapidly developing ecosystem of state-of-the-art packages that could be easily integrated with this package. Examples relevant to those in the auditory modeling field including DifferentialEquations.jl, DynamicalSystems.jl, Flux.jl, and Turing.jl.

Installation

Installation is very easy. Simply open your package REPL by pressing ] while your Julia REPL is open, and then type:

add ZilanyBruceCarney2014

The Julia package manager should then automatically install the Julia source code and download the appropriate packaged C binaries for your platform. Binaries are available for most processor architectures on Linux (using either glibc or musl), for 32- and 64-bit Windows platforms, and for 64-bit Mac platforms.

In your own scripts or packages, you can put

using ZilanyBruceCarney2014

at the top and then use the functions described below to invoke the model.

If you run into any problems, please reach out to [email protected].

Usage

There are three key functions for users of the model: sim_ihc_zbc2014, sim_anrate_zbc2014, and sim_spikes_zbc2014. You can find documentation for these functions in the source code or in the REPL, accessible by first typing ? in the Julia REPL to access help and then typing the name of the function. (Note that the package must be loaded via using before you look for help — see above for details.) Each of these functions accepts two positional arguments:

1. Vector-valued input waveform. For sim_ihc_zbc2014, this is the acoustic waveform; for the other functions, this is the output of sim_ihc_zbc2014.
2. Scalar-valued CF in Hz.

Other parameters are passed as keyword arguments. For example, simulating an IHC response at 4 kHz CF using the cat model would be done as follows:

ihc_output = sim_ihc_zbc2014(stimulus, 4e3; species="cat")

Direct bindings are available in the form of the functions IHCAN!, Synapse!, and SingleAN!, which emulate the behaviors of the corresponding C functions in the model source code. Note that the exclamation marks indicate that these functions operate on (some of) their arguments in-place, just as the original functions in C do. Most users will not need to interact with these functions.

Testing

Many basic response properties of the auditory-nerve simulations (e.g., responses grow in response in increasing sound level) are tested in test/runtests.jl. If you want to run these tests yourself, follow these steps:

• Clone the repository (git clone [email protected]:guestdaniel/ZilanyBruceCarney2014.jl)
• Change directory to the repository
• Open a Julia REPL
• Switch to the Pkg REPL (press ] on your keyboard)
• Instantiate the package's dependencies (instantiate in the Pkg REPL)
• Call test from the Pkg REPL

Funding

Development of this package was supported by the following funding resources at various points in time:

• NIH R01 DC005216
• NIH F31 DC019247
• NIH R01 DC010813
• NIH F32 DC022143
• UMN College of Liberal Arts Graduate Fellowship

License and acknowledgments

This repository is licensed under the GNU AGPL v3 license. The underlying model code is largely the work of:

• Muhammad S. A. Zilany
• Ian C. Bruce
• Laurel H. Carney