Implementation of examples from D.R. Ferreira, A Primer on Process Mining: Practical Skills with Python and Graphviz, Springer, 2017
We are going to install Jupyter lab under Python 3.6 locally.
First, go to the link Python 3.6 install and download the corresponding version of Python 3.6 of your operating system (MacOs, Linux, or Windows) and install it. Frequently, some version of Python is pre-installed in your system, but a different one version of Python 3.6.
You can find the just installed Python 3.6 with the following command in the local Terminal
$ which python3
Let us suppose, Python 3.6 is in
71EE
stalled in /usr/bin/python3. Now, we will create a virtual environment, where a set of packages will be installed and executed independently of packages installed in the local system. virtualenv is an excellent way to ensure running Jupyter lab and your particular applications correctly.
Now, we will create a virtual environment with Python 3.6. The following instruction will create a fold with the name env_py3.6, where virtualenv will install some basic Python packages. If you do not have installed virtualenv, then install it by following instructions in virvtualenv installation. Let us note that /usr/bin/python3 is the location of Python 3.6 obtained from the instruction $ which python. In addition, let us note that the folder env_py3.6 is not necessarily the folder of your data! It is just a folder with right packages for execution of Jupyter lab. Creating the virtual environment:
$ virtualenv --python=/usr/bin/python3 env_py3.6
Then, we will activate the virtual environment named env_py3.6. This will activate a prompt (env_py3.6).
(env_py3.6) $ source env_py3.6/bin/activate
Now, we will install Jupyter lab in the environment (env_py3.6):
(env_py3.6) $ pip install jupyterlab
It's done! Optionally, you can install more packages, for instance
(env_py3.6) $ pip install matplotlib
(env_py3.6) $ pip install graphviz
(env_py3.6) $ pip install networkx
(env_py3.6) $ pip install scipy
If you want, you can generate output suitable for requirements file with all installed packages
(env_py3.6) $ pip freeze
Optionally, generate a requirements file and then install it in another environment:
(env_py3.6) $ pip freeze > requirements.txt
The second form for installing is simpler and it uses requirements.txt, which you can download from this GitHub:
(env_py3.6) $ pip install -r requirements.txt
Finally, run jupyter lab in the environment (env_py3.6):
(env_py3.6) $ jupyter lab
Then, the local web page of jupyter lab will appear automatically (after some seconds) on your web browser, for example: http://localhost:8889/lab
For log out from the Jupyter lab running, press Ctrl-C in the Terminal, then log out from the environment env_py3.6 with the following instruction:
(env_py3.6) $ deactivate
For erasing the environment:
$ rm -rf env_py3.6
Simply, in your Terminal write:
$ source /Users/laengle-uchile/env_py3.6/bin/activate
Where /Users/laengle-uchile/env_py3.6/ is the global location of the virtual environment env_py3.6. Then write:
(env_py3.6) $ jupyter lab
The video contains an animated summary of the installation procedure of Jupyter lab.