How co configure vscode to work with numpy source code
I want to contribute to numpy but when I open the C I get alot of errors the most notable one it doesn't know what Python.h is not detected I am looking at the contributing documentation but there doesn't seem to be anything about setting up an Text editor or IDE there is the section Setting up and using your development environment but as I understand this section is just about how to setup testing for numpy not about my problem what do I need to setup
Hi Rudolf, This is what I do typically for SciPy on Windows, but NumPy is similar. First open the NumPy folder then get into this part to adjust your toolchain configurations. [image: image.png] Specifically the include files are needed, and fill them in properly for the project. [image: image.png] The first line is by default and will search recursively, the second line is where python.h lives for me, the third one is needed in one of the SciPy workflows so not related. There is also the json version of this if you don't like the clicky-clicky way. Then as new errors pop-up handle them one by one. Note that an IDE needs extra work to understand stuff and each machine is a bit different. Typically it is better to try to compile NumPy first and make sure necessary things are already in place and you are not wasting your time on things that does not exist. Then try to replicate it in your IDE. I recently started VS Code just to understand what the cool kids are doing so I might not be giving you the latest/greatest advice but so far it is working for me. Hope to see your contributions soon ilhan On Mon, Sep 9, 2024 at 4:26 PM Rudolf Vrbensky <ertyperty24@gmail.com> wrote:
Hi Rudolf, This is what I do typically for SciPy on Windows, but NumPy is similar. First open the NumPy folder then get into this part to adjust your toolchain configurations. [image: image.png] Specifically the include files are needed, and fill them in properly for the project. [image: image.png] The first line is by default and will search recursively, the second line is where python.h lives for me, the third one is needed in one of the SciPy workflows so not related. There is also the json version of this if you don't like the clicky-clicky way. Then as new errors pop-up handle them one by one. Note that an IDE needs extra work to understand stuff and each machine is a bit different. Typically it is better to try to compile NumPy first and make sure necessary things are already in place and you are not wasting your time on things that does not exist. Then try to replicate it in your IDE. I recently started VS Code just to understand what the cool kids are doing so I might not be giving you the latest/greatest advice but so far it is working for me. Hope to see your contributions soon ilhan On Mon, Sep 9, 2024 at 4:26 PM Rudolf Vrbensky <ertyperty24@gmail.com> wrote:
participants (2)
-
Ilhan Polat -
Rudolf Vrbensky