3 ms·
* matplotlib is wrapped by PyPlot.jl. * TensorFlow is wrapped by TensorFlow.jl, improves on Python API (https://github.com/malmaud/TensorFlow.jl/blob/master/do
by thebooktocome 9y ago
* matplotlib is wrapped by PyPlot.jl.
* TensorFlow is wrapped by TensorFlow.jl, improves on Python API (https://github.com/malmaud/TensorFlow.jl/blob/master/docs/src/why_julia.md https://github.com/malmaud/TensorFlow.jl/blob/master/docs/sr...)
* Pandas is wrapped by Pandas.jl, interfaces with Julia's DataFrame, generally excellent.
* Images.jl is a 60%-to-70% replacement for scikit-image.
* NumPy and SciPy are mostly covered by StatsBase and Base, can't think of anything off-hand that's missing.
Julia's still immature in the machine learning domain, but a lot of new work is being done. One interesting somewhat-new package is Madeleine Udell's LowRankModels.jl, which has a particularly Julian approach to low rank fitting.