Decoder models API

fit_decoder(
    fit_results::Dict, datasets::Vector, decode_vars::Vector, t_val::Vector, λ_dict::Dict;
    traces_use::Union{Nothing, Dict}=nothing, v_condition::String="all"
)

Fit a decoder to predict the values of decode_vars from the neural traces in fit_results using L1-regularized linear regression.

Arguments

• fit_results::Dict: A dictionary containing the results of fitting CePNEM to the neural data.
• datasets::Vector: A vector of dataset names to fit the decoder to.
• decode_vars::Vector: A vector of behavior variable names to decode.
• t_val::Vector: A vector of timepoints to use for cross-validation.
• λ_dict::Dict: A dictionary of regularization strengths for each variable.

Optional Arguments

• traces_use::Union{Nothing, Dict}=nothing: A dictionary of trace arrays to use instead of the ones in fit_results.
• v_condition::String="all": The velocity condition to use for selecting timepoints. Can be "all", "fwd", or "rev".

Returns

• dict_sol::Dict: A dictionary of the fitted decoder models.
• dict_pred::Dict: A dictionary of the predicted values of decode_vars for each timepoint.
• dict_cost::Dict: A dictionary of the mean squared error between the predicted and actual values of decode_vars for each timepoint.

compute_variance_explained(dict_cost::Dict, fit_results::Dict, t_val::Vector, decode_vars::Vector, λ_dict::Dict; min_P_variance::Float64=0.5)

Compute the variance explained by the decoder for each dataset, behavior variable, regularization strength, and timepoint.

Arguments

• dict_cost::Dict: A dictionary of the mean squared error between the predicted and actual values of decode_vars for each timepoint.
• fit_results::Dict: A dictionary containing the results of fitting CePNEM to the neural data.
• t_val::Vector: A vector of timepoints to use for cross-validation.
• decode_vars::Vector: A vector of behavior variable names to decode.
• λ_dict::Dict: A dictionary of regularization strengths for each variable.

Optional Arguments

• min_P_variance::Float64=0.5: The minimum variance of the P variable for a dataset to be included in the analysis.

Returns

• dict_quality::Dict: A dictionary of the variance explained by the decoder for each dataset, behavior variable, regularization strength, and timepoint.

average_dict_qualities(dict_qualities::Dict)

Compute the average variance explained by the decoder for each dataset, behavior variable, regularization strength, and timepoint.

Arguments

• dict_qualities::Dict: A dictionary of the variance explained by the decoder for each dataset, behavior variable, regularization strength, and timepoint.

Returns

• dict_quality::Dict: A dictionary of the average variance explained by the decoder for each dataset, behavior variable, regularization strength, and timepoint.