Miniconda3: No packages found in current win-32 ch

I've been following a guide in how to install Miniconda3 and use that to install different packages as NumPy and SciPy.

However, after installing Miniconda3 and opened a new project in PyCharm, the guide tells me that I should write:

conda install numpy

...which returned:

Fetching package metadata: Could not connect to https://repo.continuum.io/pkgs/free/noarch/
Could not connect to https://repo.continuum.io/pkgs/pro/win-32/
Could not connect to https://repo.continuum.io/pkgs/free/win-32/
.Could not connect to https://repo.continuum.io/pkgs/pro/noarch/
...
Error: No packages found in current win-32 channels matching: numpy

You can search for this package on Binstar with

binstar search -t conda numpy

You may need to install the Binstar command line client with

conda install binstar

Which opted me to try:

conda install binstar

...which gave me the same error (even including the last line opting me to try conda install binstar).

I also tried, as a reference, to use conda update conda which resulted in Error: No packages found in current win-32 channels matching: conda.

I've added Miniconda when installing into my PATH variables. Does anybody have any clue regarding this?

EDIT: I have 32 bit system, and I've installed the 32 bit version of Miniconda.

It looks like conda is not able to reach the package repository, so you likely have connection issues (your firewall might be blocking conda, or maybe you are behind a proxy).

Here are 3 workaround (if you cannot fix your network problem):

1. Install Anaconda instead of Miniconda

You could install Anaconda from continuum website. Unlike Miniconda, it comes with a lot of pre-installed packages such as numpy and scipy.

2. Install `numpy` from a pre-built conda archive

You can manually download the archives package from the conda repository, and install them using:

conda install numpy.tar.bz2

You need to download the archive for your operating system. Depending on the package you chose, you may have issue with dependencies...

3. Download pre-built wheel files

This site¹ provides tons of pre-built package for Windows, which are relatively easy to install:

python -m pip install numpy.whl

You should use the python executable that comes with conda (should not be an issue if you do not have multiple python installations on your system).

_{¹ I am not the author nor the maintainer of this site, I just find it very useful when you want to install python packages on Windows and cannot use pip/conda.}