How abolim/mental-health-in-tech-industry works
Project type and purpose
• This is an Exploratory Data Analysis (EDA) research project focused on mental health in the tech industry.
• The core goal is to ingest a public survey dataset, perform data wrangling and visual analysis in a Jupyter notebook, and produce output artifacts (charts, PPT slides).
File structure
• Top‑level directories:
• 1-Data: raw CSV data
• 2-Analysis: Jupyter Notebook (.ipynb) and presentation (.pptx) containing code, analysis, visuals, insights
• 3-Outputs: generated static graphics (PNG/SVG) named by research question
• Support files at root: README.md (project description), LICENSE
• No multi‑service or API tiers—this is a single‑pipeline data workflow.
README for additional insights
• Confirms use of Python data stack: pandas, numpy, seaborn, matplotlib, plotly
• Outlines research questions as discrete analysis modules
• Indicates interactive vs. static visualizations; notes pivot tables, bar plots, funnel charts
a. Main components
1. Data Ingestion (1-Data/OSMI_2019_MentalHealth_in_Tech_Survey_Results.csv)
2. Data Processing & Analysis (Jupyter Notebook in 2-Analysis)
3. Visualization Engine (matplotlib/seaborn/plotly within notebook)
4. Output Artifacts (3-Outputs: image files & PPTX)
5. External Data Source (OSMI survey dataset/Kaggle)
b. Relationships & data flow
• Raw CSV → Notebook (data cleaning & transformation) → analysis modules per research question → visualization routines → output files & presentation
• Notebook depends on Python environment and libraries (shown as a sub‑component)
c. Architectural patterns / design principles
• Modular analysis: each research question implemented as a logical block in the notebook, outputs segregated by name
• Reproducibility: single‑source notebook orchestrates entire pipeline
d. Key technologies & frameworks
• Python 3.x environment
• pandas, numpy for data manipulation
• seaborn, matplotlib, plotly for charts
• Jupyter Notebook as orchestration layer
• ORCA (plotly‑orca) for exporting static images
Detail without over‑splitting
• Logical stages: ingestion, cleaning, EDA blocks (each research question), visualization, export
• If desired, break out the five major research‑question modules as sub‑steps under “Notebook Analysis” to show modularity