# Installation¶

This section describes how to install Python datatable on various systems.

## Prerequisites¶

Python 3.5 or newer is a prerequisite. You can check your python version via

$python --version  If you don’t have Python 3.5 or later, you may want to download and install the newest version of Python, and then create and activate a virtual environment for that Python. For example: $ virtualenv --python=python3.6 ~/py36


## Install on Windows¶

Currently datatable does not work on Windows. There is an open issue #1114 to add support for Windows platforms, and there is a certain amount of progress in that direction; however, there are still some unresolved problems.

## Build from Source¶

In order to install the latest development version of datatable directly from GitHub, run the following command:

$pip install git+https://github.com/h2oai/datatable  Since datatable is written mostly in C++, you will need to have a C++ compiler on your computer. We recommend either Clang 4+, or gcc 5+, however in theory any compiler that supports C++11 should work. It is also possible to build datatable with gcc 4.8, which has only partial support of C++11 features. In this case, datatable’s functionality will be limited, and any function using regular expressions will not be supported. ## Build modified datatable¶ If you want to tweak certain features of datatable, or even add your own functionality, you are welcome to do so. 1. First, clone datatable repository from GitHub: $ git clone https://github.com/h2oai/datatable

1. Make datatable:
$make test_install$ make

1. Additional commands you may find occasionally interesting:
# Build a debug version of datatable (for example suitable for gdb debugging)
$make debug # Generate code coverage report$ make coverage

# Build a debug version of datatable using an auto-generated makefile.
# This does not work on all systems, but when it does it will work
# much faster than standard "make debug".