KuMo: Towards Cloud Morphing

Table of contents

• Visualization Tools
• Reproducing the experimental results of any related paper

Installing KuMo

In a Julia REPL (command-line interface for julia), please use the following code snippet to install KuMo.

julia using Pkg Pkg.add(url="https://github.com/iijlab/KuMo.jl")

Visualization Tools

We provide an interface to two popular visualization tools in the Julia ecosystem:

• (GL)Makie.jl for an interactive plot analysis
• (Stats)Plots.jl for generating figures in LaTeX/PDF fashion

We recommend using the Makie interface first when designing or analyzing a scenario.

Pseudo costs selection and manipulation

Using efficiently our Cloud Morphing system, KuMo, is mainly done by selecting appropriate pseudo costs function for each resource. We implemented a small tool available within KuMo to help users with standard and usual variants of monotonic and convex pseudo costs.

In the julia REPL (command-line interface for julia), please use the following code snippet.

```julia using KuMo, GLMakie

showpseudocosts() ```

Interactive plot analysis with Makie

With our Makie interface, we provide a simple way to visualize the results of a scenario. The following code snippet will generate a plot of a default scenario.

```julia using KuMo, GLMakie

show_simulation() ```

Note that by default, show_simulation() will use a simple four convex nodes' scenario available in a small scenario collection called SCENARII.

At the time of writing, the user can try out-of-the-box the following scenario through

```julia showsimulation(SCENARII[:fournodes]) # default scenario for show_simulation()

showsimulation(SCENARII[:fournodesfourusers])

show_simulation(SCENARII[:square]) ```

Generating quality plots with Plots

Users can generate high quality plots with the goal of a LaTeX formatted PDF output in the following fashion.

```julia using KuMo, StatsPlots

scenario = SCENARII[:four_nodes]

simulateandplot!(scenario; plot_type = :all, target=:all) ```

The plot_type argument can be either :all or :plot or :areaplot. The target argument can be either :all or :nodes or :links.

Please read the documentation (WIP) for more information.

Reproducing the experimental results of any related paper

Installing Julia

We recommend any user, specially if unfamiliar with the Julia language, to use juliaup. Among other things, juliaup will install the latest release of Julia and add it to the path.

Do not use the julia version available on most repositories (such as apt) as it tends to not be maintained. Either install through juliaup or by downloading the binaries on the Julia language website.

Note that this package requires at least Julia 1.8 (the latest release at the time of writing).

How to use the scripts to generate the figures in the article

Clone the repository

Either clone this repository using, for instance the following command line in a shell

shell git clone https://github.com/iijlab/KuMo.jl.git

or in a Julia REPL

julia using Pkg Pkg.develop(url="https://github.com/iijlab/KuMo.jl.git")

Note that the last command will download and install the package at the following path ~/.julia/dev/KuMo.

Change directory

Please open a terminal anywhere within the cloned repository (or cd into it).

Run the script

If one has LaTeX installed, running the script is as simple as using

```shell julia scripts/main.jl

or from anywhere within the repository

julia path/to/main.jl ```

If no LaTeX engine are available, please add the --nolatex flag.

shell julia scripts/main.jl --nolatex

The output is generated in /figures.

Note that the first execution of the script will take a while as it will download and install all the dependencies.