Installation and Usage¶

This section details the installation and usage steps for multiple modes of accessing the tool and the examples. FiMDP requires Python 3.7 or above. The dependency on external packages is minimal and they are mainly used for illustrations. The examples presented in interactive Jupyter notebooks help in getting started with the tool and also to analyze the performance for provided examples.

Binder¶

Binder creates custom computing environments from git repositories and deploys on the cloud allowing access to interactive notebooks over any web browser. Use the following steps to get started with Binder:

Click on the following link and wait for the environment to load.
once the cloud instance of Jupyter notebook begins, navigate to the examples directory and access any notebook of interest.
For detailed description of all the example notebooks, please visit the examples section Examples.

Note

Large jobs might take significant computation time on Binder as the performance is usually lower than a modern local workstation.

Docker¶

The docker image with FiMDP is published on Docker Hub. To access the package using docker, download and install docker on your machine. The default behavior of this image is to run Jupyter lab, and that is also the intended usage. To open the Jupyter lab environment in your browser, you need the following two steps.

Access the Jupyter notebooks¶

To open the interactive Jupyter notebooks with examples via Jupyter lab, open a CLI and run:

sudo docker run --rm=true -p 7777:8888 xblahoud/fimdp:cav2020

Note that the -p 7777:8888 redirects the port 8888 of the container to the port 7777 of your computer. If the latter is already used on your computer, use another number. After running the above command, access the following url in a browser in your machine:

http://localhost:7777/lab

To get started, right-click on the README.md file in the left panel and select open with > Markdown preview. If you prefer the classic Jupyter notebook environment to Jupyter lab, type tree instead of lab.

Run bash in this container¶

Open a CLI and run:

sudo docker run -it xblahoud/fimdp:cav2020 /bin/bash

and the directory contains the all the source files of the package.

Conda Installation¶

We assume that you are familiar with the Anaconda eco-system and the conda environment and have an active installation of Anaconda or Miniconda on your computer. To use our tool with the help of conda:

Create a new conda environment with the name fimdp using the following command:
```
conda create -n fimdp python=3.7
```
Clone our GitHub repository and install the required packages in the newly created environment using the following command:
```
conda install --name fimdp -c conda-forge --file requirements.txt
```
Activate the environment using the following command:
```
conda activate fimdp
```
Launch Jupyter notebook server using the following command:
```
jupyter notebook
```
Navigate local instance of Jupyter to access the examples subdirectory and access the notebooks.

Certain examples include visualizations that need the GraphViz package installed and configured. Download and install the appropriate version of the package and add the dot file to the system PATH to successfully run certain examples. If you install GraphViz package using Anaconda, make sure that to add the PATH of the dot file from Anaconda library to your system PATH.

Pip Installation¶

We assume that you have the default Python environment already configured on your computer and you intend to use our tool inside of it. If you want to create and work with Python virtual environments, please follow instructions on virtual environments.

To get the latest version of FiMDP, you can clone the GitHub repository and install the dependencies with pip3:

git clone https://github.com/xblahoud/FiMDP
cd FiMDP
pip3 install -e .

Further, certain examples include visualizations that need the GraphViz package installed and configured. Download and install the appropriate version of the package and add the dot file to the system PATH to successfully run certain examples.

After the installation, you can start a local instance of Jupyter notebook and access the examples.