gilleswaeber / dstses

Data Science in Techno-Socio-Economic Systems

Data Science in Techno-Socio-Economic Systems

The project features mass-data accumulation through an IoT Sensor, Warehousing and Processing. The future idea of usage is to integrate the measured data with additional datasets.

Directory Structure

Directory	Description
docker	Docker container used as virtual development environment
docs	Various documentations related to components/plans
hist_data	Historical data download/curation for having training data for prediction projects
opendata_converter	Converter used to convert hist_data to better format
papers	Various Papers related to prediction models
sensor	The sensor software project used to acquire sensor data
src	The source code of the overall prediction software
models	The trained models
workbooks	Various workbooks with experiments

Installation

The installation is prepared for Linux systems in anaconda.sh. If you have an Windows computer just use Anaconda, download the installer from anaconda.com[anaconda website].

Warning

Currently, installing RPy2 failes under the Linux environment (others were not tested). Do not try! It is not worth your time, since conda makes faults in the installation procedure. Furthermore, changing python versions, as suggested by the tool, does not help your case, even though it looks promising at first. Further, investigation is already underway with Nico.

Running

conda activate stes
jupyter notebook

The browser should start and present you the notebooks, in the directory your currently in, and let’s you select one.

Tip	Select under kbd:[kernel] → kbd:[change kernel] → kbd:[stes], this activates the user installed kernel, which ensures that you definitely are in the venv while you execute stuff in the notebook. If you do not see that kernel, check whether your in the venv (stes).

Documentation

You can find the documentation in the /docs directory.