You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
@@ -90,7 +90,7 @@ The reason for this low-dimensional choice is to spend most of the time in the g
The data consists of $50$ independent sequences of length $100$ each, with a number of states varying from $2$ to $10$, to which we apply all inference algorithms (with Baum-Welch performing $5$ iterations).
All benchmarks were run from Julia with `BenchmarkTools.jl`[@chenRobustBenchmarkingNoisy2016], calling Python with `PythonCall.jl`[@rowleyPythonCallJlPython2022], and plotting results with `CairoMakie.jl`[@danischMakieJlFlexible2021].
The comparison code imports `HiddenMarkovModels.jl` version 0.5.0, and it is accessible in the `libs/HMMComparison/` subfolder of our GitHub repository.
The comparison code imports `HiddenMarkovModels.jl` version 0.5.0 (commit [9e0b7ab](https://github.com/gdalle/HiddenMarkovModels.jl/commit/9e0b7ab955866523551efbc42813adc4d2b1e3dd)), and it is accessible in the [`libs/HMMComparison/`](https://github.com/gdalle/HiddenMarkovModels.jl/tree/9e0b7ab955866523551efbc42813adc4d2b1e3dd/libs/HMMComparison) subfolder of our GitHub repository.
We tried to minimize parallelism effects by running everything on a single thread, and made the assumption that the Julia-to-Python overhead is negligible compared to the algorithm runtime.
