Install getML on Linux

This installation guide explains all necessary steps to install getML on Linux. To download the getML suite, go to

https://getml.com/product

and click the download button. This will download a tarball containing everything you need to use getML: The getML engine, the getML monitor, and the Python API.

System requirements

Your Linux should meet at least the following requirements to successfully install getML

  • GLIBC 2.17 or above (check by using ldd --version)

  • If you are using Fedora 30, you need libxcrypt-compat. Install using yum install libxcrypt-compat.

  • Python 3.5 or above must be installed on your machine. Furthermore, numpy and pandas are required dependencies for the getML Python API.


Install and run the getML engine and monitor

The getML engine is the C++ backend of getML. It comes with a graphical user interface - the getML monitor - that runs in your browser. To install these components, do the following:

  1. Extract the tarball using tar -xzvf getml-VERSION-linux.tar.gz.

  2. cd into the resulting folder. Then, type ./getML. This will create a hidden folder called .getML in your home directory. If your computer has a desktop environment, you will also have a getML icon in your Applications menu.

  3. (optional) You can now install get getML command-line interface (getML CLI). See below for further instructions.

  4. Open a browser and visit http://localhost:1709/ (if launching getML did not point you there automatically). You will see the login screen below. When installing getML for the first time select ‘Create a new account’.

  5. You will be forwarded to a registration form. Fill all required fields and hit the check button at the bottom to create a new account. By doing so you accept our terms of service, which can be found here.

../../_images/monitor_screen1.png
  1. Confirm your account using the confirmation e-mail that has been sent to you. Check the e-mail address you provided during registration and hit the orange button. Now, you should be able to log into the getML monitor. If this is not the case, please contact our support.

Install the getML Python API

The Python API is a convenient way to interact with and to control the getML engine. There are two options to install the getML Python API

From source

The source code of the Python API is shipped with the bundle downloaded too. Double-click it again and copy the Python folder to any location on your computer. Open a terminal, cd into that folder and run

python setup.py install

Install the getML CLI

getML comes with a command-line interface (CLI) that lets you configure the most important parameters on startup. The CLI is a standalone Golang-binary located in the downloaded bundle.

Note

Before you can use the CLI, you have to follow steps 1 and 2 of the installation instructions above.

After deflating the tarball, you should find the getML binary in the resulting folder.

After you have started getML for the first time, you can move the getML binary anywhere you want. We recommend moving the getML binary to a location included in the PATH environment variable, such as ~/.local/bin. You can inspect the content of the aforementioned variable in a shell using

echo $PATH

and check if it can be properly found by executing

which getML

If this returns the location you moved the binary to, you are ready to go.

For further help on how to use the CLI, just use getML -h or getML -help:

Uninstall getML

To uninstall getML from your computer

  1. Remove the folder .getML from your home directory. To do so, open a terminal and enter the following command

rm -r $HOME/.getML
  1. Delete getML binary from wherever you have put it (if you have decided to install the getML CLI).

Where to go next

The Getting started guide provides an overview of the functionality of getML and a basic example of how to use the Python API. In order to get help or provide feedback, please contact our support.