An archive of earlier notebook and SQL exercises completed while building foundations in data analysis, web scraping, machine learning and relational querying.
This repository is intentionally labelled legacy learning work. It is not part of the current flagship portfolio, and some notebooks depend on Google Drive paths, older websites or datasets that are not packaged reproducibly.
| Area | Contents | Current value |
|---|---|---|
German_car Price_Analysis/ |
Data cleaning, exploratory analysis and price-importance/prediction notebooks | Strongest earlier analysis case; worth reviewing before any standalone rewrite |
SQL_projects/ |
Danny's Diner case study and a data-wrangling exercise | SQL practice; needs dialect cleanup, validation and individual documentation |
WEB_SCRAPPING_E_COMMERCE_SITE_(EBAY)_.ipynb |
Requests/BeautifulSoup extraction and exploratory analysis | Historical scraping exercise; live page structure may have changed |
AI_NLP_01_2.ipynb |
Tokenisation, stop-word removal, stemming and lemmatisation practice | Introductory NLP notebook; duplicated in another repository |
projects.lnk |
Local shortcut file | Not portable and should be removed in a future cleanup |
- Python and pandas data cleaning;
- exploratory visualisation with Matplotlib and Seaborn;
- introductory machine-learning workflows;
- HTTP requests and HTML parsing with BeautifulSoup;
- SQL joins, aggregation, CTEs, window functions and conditional logic;
- introductory NLTK text processing.
The current target profile is Business Intelligence and Data Engineering. For that reason, this repository should remain an archive until one component is rebuilt to current standards.
The German car analysis is the best candidate for rescue because it can connect:
- data quality and feature preparation;
- analytical explanation;
- German-market context;
- model limitations and reproducibility.
The SQL exercises are useful practice but are weaker than the dedicated SQL warehouse repository.
- the root README previously described only one dataset and did not explain the mixed folder structure;
- notebook links reference old repository names and Google Drive paths;
- the NLP notebook is duplicated in
Coach_bot; - SQL files mix styles/dialects and include unverified query logic;
- there is no environment file, test suite or reproducible execution order;
- output-heavy notebooks make the repository large.
- Choose one project and move it into a dedicated repository.
- Package a safe sample dataset or document its public source.
- Remove personal/local paths and stale links.
- Clear unnecessary notebook outputs.
- Add problem, method, validation, findings, limitations and run instructions.
- Add a small test or data-quality section.
- Leave the remaining exercises here as archive material.
Archive — not recommended for profile pinning.