# ATLASx Analysis tools 

The data and scripts contained in this repository allow the user to reproduce the figures in the ATLASx manuscript from the original data files.

## Installation

The installation can be completed in less than 10 minutes, including installation of dependencies and fetching the data from the git repository.

### Requirements

- python 2.7 or higher (Tested and recommended: Python 3.9.6)
- pip

The code is adapted for python 3, but it can also be executed in python 2.s

Runtimes are indicated for each script and were determined normal desktop computer (macOS).

### Download repository

`$ git clone https://c4science.ch/source/ATLASxAnalysis.git`

To install the required dependencies: 

`$ cd ATLASxAnalysis`
`$ make`

### Note
Data files are stored using git large file storage (lfs). The make file will install git lfs automatically. However, if lfs was not installed previously, the repository has to be updated after installation:

`$ git pull`

This is needed to retrieve the data files from the repository after installation.

## Usage

### Network Analysis

`$ cd NetworkAnalysis/Source`

Plot component distribution for database scopes

`$ python3 get_component_distribution.py #Runtime: 97s`

By default, the database scopes will be plotted. For a resolution by data source, add data_sources as an argument to the above command:

`$ python3 get_component_distribution.py data_sources #Runtime: 11s`

### CSV file conversion
CVS files of networks are quite practical for visualisation, e.g. in the Gephi software.
To convert the gpickle files for database scopes to CSV, run the following:

`$ python3 print_csv_from_gpickle.py #Runtime: 52s`

As above, for single data sources use:

`$ python3 print_csv_from_gpickle.py data_sources #Runtime: 50s`

The output is automatically written to a new folder called ATLASxAnalyses_output created in the same directory as the repository.

The data files used in the repository is the same as the one used in the manuscript, which has been downloaded on 8 November 2020.
For updated network files, please contact the authors of the paper directly.