MnemonicDiscriminationIndex.jl

Package for our new index of mnemonic discrimination

To install this package to you environment, run

julia> ]add MnemonicDiscriminationIndex.jl

in the Julia REPL. To prevent breakage, add a compat entry to the version you install in your environment's Project.toml as the API might be improved over time.

For a workflow similar to the paper using the logistic5, lets assume we have our MST data.

julia> using MnemonicDiscriminationIndex

julia> old_or_new = [0,0,0,1,0,1,1,1]
8-element Vector{Int64}:
 0
 0
 0
 1
 0
 1
 1
 1

julia> dissimilarity = 0:(1/7):1
0.0:0.14285714285714285:1.0

Then, we fit our data to the logistic5 curve (using a reproducible rng) and save the parameters:

julia> using StableRNGs

julia> logistic5_results = fit_logistic5(dissimilarity, old_or_new; rng=StableRNG(123));

Now, we have an MDIResult object with all the information needed:

julia> logistic5_results.auc
0.44583208733243396

julia> logistic5_results.Δ
0.9390919680940668

julia> logistic5_results.λ
0.5252519428557438

To fit to a different function than logistic5, check out the docstrings for fit_model by calling ?fit_model in the REPL.

Code for the paper can be found here.

Release v0.1.0 of this repository is the version of the code used in the paper.