Installation instructions for the ParaMonte Python library | ParaMonte: Parallel Monte Carlo Library

Tip: The Python distribution that we recommend you to use is Anaconda Python 3. Anaconda Python automatically ships with all the external module dependencies of the ParaMonte Python library, which are not that many (e.g., numpy, scipy, pandas, matplotlib, seaborn). Once Anaconda is installed on your system, find and open an Anaconda-provided Windows command-prompt from the Windows startup, as shown in the figure below.

There are three ways to install and use the ParaMonte library from Python:

The fastest way

Install the latest release of the ParaMonte Python library via the Python Package Index (PyPI). To do so,

Open an Anaconda command-prompt (on Windows) or a Bash shell (on Unix-like OS such as Linux or Darwin/macOS), then type,
```
pip install --user --upgrade paramonte
```
This will install the Python interface to ParaMonte on your system along with any dependencies. If you are inside an ipython or jupyter session, try instead,
```
!pip install --user --upgrade paramonte
```
Then, open a Python (or ipython or jupyter) session and type,
```
import paramonte as pm
```
The ParaMonte Python library is ready to use!

The fast way

Download and unzip the prebuilt Python package for ParaMonte from the project repository’s release page on GitHub:
- for Windows: libparamonte_python_windows_x64.zip.
- for Linux: libparamonte_python_linux_x64.tar.gz.
- for macOS: libparamonte_python_darwin_x64.tar.gz.
  Important: If you are a macOS user, we strongly recommend you to NOT download the library directly from the the library's GitHub release page and NOT by clicking directly on the above link. Instead, use the curl command line program to download your the library from the release page. To do so, first make sure you have curl installed on your system. If it is not installed, you can install it via Homebrew,
```
brew install curl
```
  Then, copy the web-link to the ParaMonte Python library and download it via curl to the current folder. For example,
```
libname=libparamonte_python_darwin_x64
curl -OL https://github.com/cdslaborg/paramonte/releases/download/v1.5.1/$libname.tar.gz
tar xvzf $libname.tar.gz && cd $libname
```
Open an Anaconda command-prompt (on Windows) or a Bash shell (on Unix-like OS such as Linux or Darwin/macOS).
Change the working directory of your shell environment to the unzipped directory of the ParaMonte Python library, such that when you list the contents of your current working directory, you see a subdirectory named paramonte. This subdirectory contains the Python module interface to the ParaMonte library.
Open an ipython or jupyter session in this environment, and type import paramonte as pm.
Congratulations! you have successfully installed the Python interface to the ParaMonte library on your system.

Important: Since this is only a local manual installation of the ParaMonte Python library (as opposed to the direct installation from PyPI), every time that you intend to use the ParaMonte library, you will have to navigate to the parent directory containing the paramonte subdirectory. Otherwise, the Python interpreter will not be able to find and import paramonte.

Important: If you are a macOS (Darwin) user and you have downloaded the prebuilt ParaMonte libraries directly via a web-browser from the GitHub release page to use on your macOS system, read this troubleshooting page before attempting to run ParaMonte simulations.

The long way

If your platform is other than Windows/Linux/macOS or, if your system’s architecture is other than amd64, also known as x64, or, if you are a professional programmer constantly challenging your computer programming knowledge and skills, then the last resort to install the ParaMonte Python library on your system is to build the library from scratch on your system. In such a case, you will have to follow the special instructions given for your platform on this website,

Note: On some platforms (e.g., supercomputers) the support for Python’s visualization libraries such as matplotlib and seaborn may be weak. In particular, import matplotlib is known to cause a segmentation fault error on some platforms, which subsequently leads to the crash of the active Python session. The Python interface of ParaMonte has been implemented in such a way as to ensure that no unnecessary call is made to the visualization libraries as long as the user is only interested in performing a pure ParaMonte simulation, without any visualizations or post-processing of the current (or some old preexisting) ParaMonte simulation output files.

If you have any questions about the topics discussed on this page, feel free to ask in the comment section below, or raise an issue on the GitHub page of the library, or reach out to the ParaMonte library authors.