How to install NumPy for Python 3.6

Question

I am using Python 3.6b3 for a long running project, developing on Windows. For this project I also need NumPy. I\'ve tried Python36 -m pip install numpy, but it seems that p

Answer 1

Pip is always be installed in subdirectory \Scripts of the Python 3.6 install directory.

Open command prompt in windows, change directory to \Scripts and input pip list, you will see what has been installed, then input pip install numpy, and the numpy will be installed.

Answer 2

As long as binary packages (so-called 'wheels') for 3.6 have not been released to PyPi yet, you can resort to unofficial (but working) ones available at http://www.lfd.uci.edu/~gohlke/pythonlibs/. Download the file and install it like this:

pip install C:\path\to\numpy‑1.11.3+mkl‑cp36‑cp36m‑win_amd64.whl

Answer 3

EDIT: I misread the question, disregard this answer.

Try manually downloading the wheel from pypi and installing manually. On windows, it should look something like this:

C:/path/to/python36/pip.exe install C:/path/to/numpy_foo_.whl

Answer 4

Pip itself is not part of Python. However, since version 3.4 Python includes the ensurepip module that allows bootstrapping a pip installation. After calling

python3.6 -m ensurepip

you should be able to install numpy via pip.

Answer 5

Here is the list of steps needed to be performed for installing numpy on windows. This worked for me.

1. Set path variable for python- C:\Users\XXXXXXX\AppData\Local\Programs\Python\Python36-32
2. Download get-pip.py locally
3. Open cmd and navigate to get-pip.py file
4. run python get-pip.py for installation of PIP
5. set pip path in env variable path - C:\Users\XXXXXXX\AppData\Local\Programs\Python\Python36-32\Scripts;
6. pip install numpy

Answer 6

You can install numpy through command prompt using pip install command. Follow the below steps: 1) In your command prompt, go to the scripts folder inside your Python 3.x installation.

2) Type following command on your command prompt

install numpy

Refer the attached screenshot showing both the steps.