0 reply: I guess you are using Windows. Although many people recommend that you use Enthought and Anaconda, it will deprive you of a lot of learning experience. If you know how to use CMD/Powershell, first put the l/ib folder under the Python folder into your system variables, in addition, pip comes with Python 2.7.9 installation. (If you don't know this, you should use Pre-installed IDE.) in cmd, using pip to install numpy in a 64-bit system may cause problems, which many people have encountered, it is estimated that you will need to install Scipy later. This requires C and Fortran compilers ......
So I suggest you install the wheel file, I will give you an address http://www.lfd.uci.edu /~ Gohlke/pythonlibs/# numpy
Make sure that you select the wheel file based on your python version and the number of bits on your computer. To install the wheel file, you must first install the package. Enter cmd
pip install wheel
What OS and version. Numpy installation is not particularly troublesome. Of course, it is difficult to build one from the source code in windows. We recommend that you install Anaconda: Anaconda Scientific Python Distribution.
Many scientific computing libraries have been integrated (including not only numpy, but also sklearn and pandas ). Python2.7 python3.4. Use Anaconda. This is the method I recorded previously to install numpy: A compilation error occurs when Python 3 installs a third-party library.
Download the zipping package (or tar.gz)
Python setup. py install you need python (x, y) Updates: Although there are few people looking at this answer, please update it to prevent anyone from being misled by me.
It seems that the website cannot be opened. Therefore, I recommend another method. Users in windows can directly install python (x, y), which contains almost all common python libraries and many practical tools, the installation success rate is high.
The following is the original answer.
............................................................
Teach you the simplest way.
Open your browser and enter datahub. top
Register an account
You can apply for an ipython notebook cloud service.
Then you can use it directly on any platform, such as computers, mobile phones, tablets, and even the old Nokia.
It has integrated all scientific computing libraries. No configuration is required. There are also detailed tutorials.
I like numpy's official website. I have the .exe installation package ~~ If you are using a windows version, you will feel a lot better if you try this website.
Http://www.lfd.uci.edu /~ Gohlke/pythonlibs/
