Stimgen Types

Abstract supertype for all stimulus generation.

BinnedStimgen <: Stimgen

Abstract supertype for all binned stimulus generation.

Bernoulli(; kwargs...) <: BinnedStimgen

Stimulus generation type in which each tonotopic bin has a probability bin_prob of being at 0 dB, otherwise it is at -100 dB.

Keywords

• min_freq::Real = 100: The minimum frequency in range from which to sample.
• max_freq::Real = 22e3: The maximum frequency in range from which to sample.
• duration::Real = 0.5: The length of time for which stimuli are played in seconds.
• Fs::Real = 44.1e3: The frequency of the stimuli in Hz.
• bin_prob::Real=0.3: The probability of a bin being filled.
• n_bins::Integer = 100: The number of bins into which to partition the frequency range.

Brimijoin(; kwargs...) <: BinnedStimgen

Stimulus generation type in which each tonotopic bin is filled with an amplitude value from an equidistant list with equal probability.

Keywords

• min_freq::Real = 100: The minimum frequency in range from which to sample.
• max_freq::Real = 22e3: The maximum frequency in range from which to sample.
• duration::Real = 0.5: The length of time for which stimuli are played in seconds.
• Fs::Real = 44.1e3: The frequency of the stimuli in Hz.
• amp_min::Real = -20: The lowest dB value a bin can have.
• amp_max::Real = 0: The highest dB value a bin can have.
• amp_step::Int = 6: The number of evenly spaced steps between amp_min and amp_max.
• n_bins::Integer = 100: The number of bins into which to partition the frequency range.

BrimijoinGaussianSmoothed(; kwargs...) <: BinnedStimgen

Stimulus generation type in which each tonotopic bin is filled by a Gaussian with a maximum amplitude value chosen from an equidistant list with equal probability.

Keywords

• min_freq::Real = 100: The minimum frequency in range from which to sample.
• max_freq::Real = 22e3: The maximum frequency in range from which to sample.
• duration::Real = 0.5: The length of time for which stimuli are played in seconds.
• Fs::Real = 44.1e3: The frequency of the stimuli in Hz.
• amp_min::Real = -20: The lowest dB value a bin can have.
• amp_max::Real = 0: The highest dB value a bin can have.
• amp_step::Int = 6: The number of evenly spaced steps between amp_min and amp_max.
• n_bins::Integer = 100: The number of bins into which to partition the frequency range.

GaussianNoise(; kwargs...) <: BinnedStimgen

Stimulus generation type in which each tonotopic bin is filled with amplitude chosen from a Gaussian distribution.

Keywords

• min_freq::Real = 100: The minimum frequency in range from which to sample.
• max_freq::Real = 22e3: The maximum frequency in range from which to sample.
• duration::Real = 0.5: The length of time for which stimuli are played in seconds.
• Fs::Real = 44.1e3: The frequency of the stimuli in Hz.
• n_bins::Integer = 100: The number of bins into which to partition the frequency range.
• amplitude_mean::Real = -10: The mean of the Gaussian.
• amplitude_var::Real = 3: The variance of the Gaussian.

GaussianPrior(; kwargs...) <: BinnedStimgen

Stimulus generation type in which the number of filled bins is selected from from a Gaussian distribution with known mean and variance parameters.

Keywords

• min_freq::Real = 100: The minimum frequency in range from which to sample.
• max_freq::Real = 22e3: The maximum frequency in range from which to sample.
• duration::Real = 0.5: The length of time for which stimuli are played in seconds.
• Fs::Real = 44.1e3: The frequency of the stimuli in Hz.
• n_bins_filled_var::Real = 1: The variance of number of bins that may be filled on any stimuli.
• n_bins_filled_mean::Integer = 20: The mean number of bins that may be filled on any stimuli.
• n_bins::Integer = 100: The number of bins into which to partition the frequency range.

PowerDistribution(; kwargs...) <: BinnedStimgen

Stimulus generation type in which the frequencies in each bin are sampled from a power distribution learned from tinnitus examples.

Keywords

• min_freq::Real = 100: The minimum frequency in range from which to sample.
• max_freq::Real = 22e3: The maximum frequency in range from which to sample.
• duration::Real = 0.5: The length of time for which stimuli are played in seconds.
• Fs::Real = 44.1e3: The frequency of the stimuli in Hz.
• n_bins::Integer = 100: The number of bins into which to partition the frequency range.
• distribution_filepath::AbstractString=joinpath(@__DIR__, "distribution.csv"): The filepath to the default power distribution from which stimuli are generated

UniformNoise(; kwargs...) <: BinnedStimgen

Stimulus generation type in which each tonotopic bin is filled with amplitude chosen from a Uniform distribution.

Keywords

• min_freq::Real = 100: The minimum frequency in range from which to sample.
• max_freq::Real = 22e3: The maximum frequency in range from which to sample.
• duration::Real = 0.5: The length of time for which stimuli are played in seconds.
• Fs::Real = 44.1e3: The frequency of the stimuli in Hz.
• n_bins::Integer = 100: The number of bins into which to partition the frequency range.

UniformPrior(; kwargs...) <: BinnedStimgen

Stimulus generation type in which the number of filled bins is selected from the Uniform distribution on the interval [min_bins, max_bins].

Keywords

• min_freq::Real = 100: The minimum frequency in range from which to sample.
• max_freq::Real = 22e3: The maximum frequency in range from which to sample.
• duration::Real = 0.5: The length of time for which stimuli are played in seconds.
• Fs::Real = 44.1e3: The frequency of the stimuli in Hz.
• min_bins::Integer = 10: The minimum number of bins that may be filled on any stimuli.
• max_bins::Integer = 50: The maximum number of bins that may be filled on any stimuli.
• n_bins::Integer = 100: The number of bins into which to partition the frequency range.

UniformPriorWeightedSampling(; kwargs...) <: BinnedStimgen

Stimulus generation type in which each tonotopic bin is filled from a uniform distribution on [min_bins, max_bins] but which bins are filled is determined by a non-uniform distribution.

Keywords

• min_freq::Real = 100: The minimum frequency in range from which to sample.
• max_freq::Real = 22e3: The maximum frequency in range from which to sample.
• duration::Real = 0.5: The length of time for which stimuli are played in seconds.
• Fs::Real = 44.1e3: The frequency of the stimuli in Hz.
• alpha_::Real = 1: The tuning parameter that exponentiates the number of unique frequencies in each bin.
• min_bins::Integer = 10: The minimum number of bins that may be filled on any stimuli.
• max_bins::Integer = 50: The maximum number of bins that may be filled on any stimuli.
• n_bins::Integer = 100: The number of bins into which to partition the frequency range.

UniformNoiseNoBins(; kwargs...) <: Stimgen

Stimulus generation type in which each frequency is chosen from a uniform distribution on [-100, 0] dB.

Keywords

• min_freq::Real = 100: The minimum frequency in range from which to sample.
• max_freq::Real = 22e3: The maximum frequency in range from which to sample.
• duration::Real = 0.5: The length of time for which stimuli are played in seconds.
• Fs::Real = 44.1e3: The frequency of the stimuli in Hz.

GaussianNoiseNoBins(; kwargs...) <: Stimgen

Stimulus generation type in which each frequency's amplitude is chosen according to a Gaussian distribution.

Keywords

• min_freq::Real = 100: The minimum frequency in range from which to sample.
• max_freq::Real = 22e3: The maximum frequency in range from which to sample.
• duration::Real = 0.5: The length of time for which stimuli are played in seconds.
• Fs::Real = 44.1e3: The frequency of the stimuli in Hz.
• amplitude_mean::Real = -10: The mean of the Gaussian.
• amplitude_var::Real = 3: The variance of the Gaussian.