PyMBAR.jl

PyMBAR.jl is a Julia package that functions as a lightweight wrapper for the Python pymbar package, facilitating the calculation of free energy differences. Leveraging PyCall.jl to call Python packages, it aims to be a drop-in replacement, making it simpler to add pymbar as a dependency in your Julia projects.

To install PyMBAR.jl, use the following command in the Julia REPL:

using Pkg
Pkg.add("PyMBAR")

Note that this requires that PyCall be able to install pymbar in the python environment. It will automatically be installed if PyCall is configured to use a conda environment. Otherwise, you will need to install pymbar in whatever python environment PyCall is configured to use.

Below, find a practical example illustrating how to utilize PyMBAR.jl. Note that all functionality available in pymbar is accessible in a similar manner.

In this example, we generate dummy data and compute the statistical inefficiency.

using PyMBAR, Random, Distributions

timeseries = PyMBAR.pymbar.timeseries

function dummy_data()
    N = 10000
    K = 1
    var = ones(N)

    for replica = 2:K
        var = vcat(var, ones(N))
    end

    X = reshape(rand(Normal(0.0, 1.0), K * N) .* var / 10.0, (K, N))
    Y = reshape(rand(Normal(0.0, 1.0), K * N) .* var, (K, N))

    energy = 10 * (X .^ 2) / 2.0 .+ (Y .^ 2) / 2.0

    return X, Y, energy
end

X, Y, energy = dummy_data()
g0 = timeseries.statistical_inefficiency(X[1, :])