Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/cknoll/sociogram

rudimentary toolkit to generate a sociogram
https://github.com/cknoll/sociogram

graph jupyter-notebook sociogram visualization

Last synced: about 18 hours ago
JSON representation

rudimentary toolkit to generate a sociogram

Awesome Lists containing this project

README 

        
# sociogram

Rudimentary toolkit to generate a sociogram.



Some future teachers have to create a [Sociogram](https://en.wikipedia.org/wiki/Sociogram) as part of their training. This repo aims to mitigate the apparently suboptimal software support for this task.



![example image](example.png)



## Data Collection and Preparation



**Ensure that the collected data is anonymized before processing it on a public web service like this!**



The data can be collected with a single question survey like: "How much do you like your classmates? Give each of the following names a rating on a scale from ⭐ to ⭐⭐⭐⭐⭐.



The results of this survey can be written to a table with N rows and N columns. Each column corresponds to one response-questionnaire. Each cell of that column containing a number between 1 and 5, except the cell corresponding to the author student themselves, which should contain zero. When ordering the columns in the same way as the names on the list the resulting table can be interpreted as a quadratic NxN matrix with zeros on its main diagonal (from upper left to lower right). This matrix should be exported as CSV-file (with some standard spreadsheet software like libre office calc). CSV is a text format with "**c**omma **s**eparated **v**alues". The next step is to open that text file in an text editor (**not** in a spreadsheet program) and to copy its complete content to the clipboard. You will paste it later (see below).



## Usage



The toolkit is implemented as jupyter notebook which can be run directly in the browser via the webservice https://mybinder.org/.



- Preview jupyter notebook: [https://nbviewer.org/...](https://nbviewer.org/github/cknoll/sociogram/blob/main/notebooks/sociogram.ipynb) (fast, but only the static example).

- Run jupyter notebook (not so fast but here you can use your own data):

    - Open this link in a new tab: [![Binder](https://mybinder.org/badge_logo.svg)](https://mybinder.org/v2/gh/cknoll/sociogram/main?urlpath=/tree/notebooks).

        - This will start a virtual machine on the remote server, install all necessary dependencies and run the notebook. Note: No software is installed on your device. Everything happens on the server provided by binder.

    - If the virtual machine is up and ready open the notebook `sociogram.ipynb`

    - Paste your CSV data (from the text editor) in the second cell

    - Chose *Cell* → *Run all* from the menu to create graphical output for your data.

    - Use *Save image as ...* from the context menu (right click on image) to store the image to your device.



## Background



The data from the table is converted to a *symmetric* matrix, which is then interpreted as the [adjacency matrix](https://en.wikipedia.org/wiki/Adjacency_matrix) of an undirected graph. This graph is then rendered with a [Force-directed graph drawing](https://en.wikipedia.org/wiki/Force-directed_graph_drawing#CITEREFFruchtermanReingold1991) based on an algorithm by Fruchterman and Reingold (1991) which is implemented in the python library `networkx` as "spring layout". To achieve better visual results (to render clusters more distinct) the small values (<=3) in the matrix are decreased by factor 10.



This visualization method was loosely inspired by the blog post: .