NumPy Illustrated: The Visual Guide to NumPy

Re:NumPy Illustrated: The Visual Guide to NumPy by fishmaster on Tuesday December 29, 2020 @09:18AM (#25569906)

This is exactly why I'm going to Julia. It either has these constructs in the language itself or is powerful enough to express them adequately. Numpy is a necessary crutch for Python.

[ Parent | on HN ]

Re:NumPy Illustrated: The Visual Guide to NumPy by pc86 on Tuesday December 29, 2020 @11:21AM (#25571304)

You're going to Julia so that you can avoid typing "import numpy as np" at the beginning of a file? Or is there more to it?

[ Parent | on HN ]

Re:NumPy Illustrated: The Visual Guide to NumPy by dklend122 on Wednesday December 30, 2020 @08:46AM (#25581596)

Mathematical programming in Julia is just more natural and integrated with the language and type system.

From dot broadcasting, to multiple dispatch, to parametric types (if you want ), multidimensional array comprehensions. No worrying about 4 different container types (list, np array, pytorch tensor, symbolic etc etc)

Everything works with one array abstraction, including gpu and multi threaded

[ Parent | on HN ]

Re:NumPy Illustrated: The Visual Guide to NumPy by fishmaster on Thursday December 31, 2020 @10:51AM (#25593723)

This is what I mean with it being more natural and elegant in Julia. Also, "@code_warntype" is incredibly useful for efficient code.

[ Parent | on HN ]

Re:NumPy Illustrated: The Visual Guide to NumPy by axi1 on Wednesday December 30, 2020 @04:07AM (#25579980)

As a sidenote, I don't write "import numpy as np" in jupyter notebooks (at least as long as I don't share them with others) for I have this line in my ipython startup file.

[ Parent | on HN ]

Re:NumPy Illustrated: The Visual Guide to NumPy by fishmaster on Tuesday December 29, 2020 @02:08PM (#25573509)

Also so that I don't have to use things like "np.dot(A, B)" and similar annoyances.

[ Parent | on HN ]