Job Market Analysis

This project matches students to Zurich region tech jobs based on their ZHAW degree, salary expectations and preferred workload percentage.

Python Environment Setup and Management

Install conda environment:

$ conda env create -f jobMarket.yml

Update the environment with new packages/versions:

1. modify template.yml
2. run conda env update:

$ conda env update --name jobMarket --file jobMarket.yml --prune

prune uninstalls dependencies which were removed from sample.yml

Use environment: before working on the project always make sure you have the environment activated:

$ conda activate jobMarket

Check the version of a specific package (e.g. html5lib) in the environment:

$ conda list scipy

Export an environment file across platforms: Include only the packages that were specifically installed. Dependencies will be resolved upon installation

$ conda env export --from-history > jobMarket.yml

List all installed environments: From the base environment run

$ conda info --envs

Remove environment:

$ conda env remove -n jobMarket

See the complete documentation on managing conda-environments.

Runtime Configuration with Environment Variables

The environment variables are specified in a .env-File, which is never commited into version control, as it may contain secrets. The repo just contains the file .env.template to demonstrate how environment variables are specified.

You have to create a local copy of .env.template in the project root folder and the easiest is to just rename it to .env.

The content of the .env-file is then read by the pypi-dependency: python-dotenv. Usage:

import os
from dotenv import load_dotenv

load_dotenv reads the .env-file and sets the environment variables:

load_dotenv()

which can then be accessed (assuming the file contains a line SAMPLE_VAR=<some value>):

os.environ['SAMPLE_VAR']

Project Organisation

According to Is It Ops That Make Data Science Scientific? Archives of Data Science, Series A, vol 8, p. 12, 2022.

Code and configurations used in the different project phases are stored in the subfolders

• data_acquisition
• eda
• modelling
• deployment

Artefacts from the different project phases are provided in the subfolder docs:

• Project charta
• Data report
• Modelling report
• Evaluation decision log

Further Information

• "About Readmes" on Github https://docs.github.com/en/repositories/managing-your-repositorys-settings-and-features/customizing-your-repository/about-readmes
• Python Dev Guide

API Keys

All API keys included in this project were burnt before publishing.