mastrof / meansquareddisplacement.jl

Evaluate MSD of timeseries

MeanSquaredDisplacement

MeanSquaredDisplacement.jl provides Julia functions (imsd and emsd) to compute the mean squared displacement (MSD) of timeseries.

imsd is used to evaluate the MSD of individual timeseries, while emsd evaluates ensemble averages.

MeanSquaredDisplacement.jl relies on Autocorrelations.jl to compute MSDs, which uses a plain correlation algorithm (O(N^2)) for short timeseries and a FFT-based algorithm (O(NlogN)) for larger ones, ensuring good performance for all sample sizes.

The MSD of non-scalar timeseries can also be evaluated (equivalent to summing the MSD of each scalar element), but it's not yet optimized.

MSD of a single timeseries

using MeanSquaredDisplacement
x = cumsum(randn(10000))
imsd(x)

Individual MSD of multiple timeseries

using MeanSquaredDisplacement
x = cumsum(randn(10000, 100))
imsd(x) # evaluates MSD along columns

MSD of uneven timeseries

using MeanSquaredDisplacement
N = 100
Tmin, Tmax = 100, 2000
# trajectories of random length between Tmin and Tmax
x = [cumsum(randn(rand(Tmin:Tmax))) for _ in 1:N]
iM = imsd.(x)
eM = emsd(x)

meansquareddisplacement.jl's Issues

TagBot trigger issue

This issue is used to trigger TagBot; feel free to unsubscribe.

If you haven't already, you should update your TagBot.yml to include issue comment triggers.
Please see this post on Discourse for instructions and more details.

If you'd like for me to do this for you, comment TagBot fix on this issue.
I'll open a PR within a few hours, please be patient!

Registration

@JuliaRegistrator register()

Reimplement autodot?

https://github.com/mastrof/MeanSquaredDisplacement.jl/blob/2ea9b566bef58ea08bed2dbd6dfc36100766bb49/src/acf.jl#L69C1-L69C53

Might be safer to reimplement _autodot locally since it is not exported by StatsBase.
Same holds for check_lags.

Move `acf` routines to Autocorrelations.jl

autocorrelation functions will be moved to Autocorrelations.jl, only unfolding and msd evaluation will be kept here