Data Analytics

Areas We Can Work Through

• Excel for Analysis: clean messy spreadsheets, apply core formulas (XLOOKUP, INDEX/MATCH, SUMIFS, IF logic), build pivot tables and pivot charts, and use Power Query for repeatable imports.
• Data Collection & Sources: pull data from CSV, Excel, Google Sheets, public datasets (Stats SA, Kaggle, World Bank, data.gov) and REST APIs, and work confidently with CSV, JSON and Excel formats.
• SQL Foundations: write SELECT, WHERE, ORDER BY and DISTINCT queries, and summarise data with aggregate functions, GROUP BY and HAVING.
• Intermediate SQL & Data Modelling: combine tables with INNER and LEFT joins, subqueries, CTEs and window functions (ROW_NUMBER, RANK, running totals), and understand keys, relationships and tidy-data principles.
• Python for Analysts: use pandas to load, explore, clean, filter, group, merge and reshape (pivot/melt) real datasets, and automate repetitive analysis.
• Data Cleaning & Preparation: handle missing values, duplicates, inconsistent categories, wrong data types, messy dates and outliers, and reconcile data from several sources.
• Exploratory Data Analysis (EDA) & Statistics for Analysts: apply descriptive statistics, distributions, correlation, trend and seasonality checks and segmentation, and avoid misleading conclusions from a sample.
• Dashboards & Visualisation: choose and label the right chart, and build interactive dashboards in Looker Studio (free, browser-based) and with Python (Plotly) for a non-technical audience.
• Reporting & Data Storytelling: turn analysis into a written report and a short presentation, with an executive summary, clearly defined KPIs and recommendations a manager can act on.
• Capstone: take a messy, multi-source dataset end-to-end, clean it, analyse it, build a dashboard, and deliver a decision-ready report with recommendations.

Areas We Can Work Through

• Python Foundations: use Python (Pandas, NumPy, SciPy) with data types, loops, functions, DataFrames, and Jupyter Notebooks to explore real-world datasets, including file handling for CSV, Excel, JSON, and databases.
• Data Collection & Sources: collect and work with data from Kaggle, GitHub, the UCI ML repo, REST APIs, SQL databases (PostgreSQL, MySQL), and web scraping with BeautifulSoup and Requests.
• Version Control with Git & GitHub: use Git fundamentals (init, add, commit, status), branching and merging, remote repos, and collaborate via GitHub to build a project portfolio.
• Data Cleaning & Wrangling: handle missing values and outliers, convert data types, filter, group, aggregate, merge and join datasets, and engineer new features.
• Exploratory Data Analysis (EDA) & Visualisation: identify trends and create compelling visualisations with Matplotlib, Seaborn, and Plotly, including histograms, scatter plots, and correlation heatmaps.
• Statistics for Data Science: apply descriptive statistics, probability distributions, and confidence intervals to draw valid insights and avoid common pitfalls.
• Supervised Learning: build regression models (Linear, Ridge, Lasso) and classification models (logistic regression, decision trees, random forest, KNN).
• Unsupervised Learning: apply clustering (K-means) and PCA for pattern discovery.
• Model Evaluation & Optimisation: use train/test splits and cross-validation, evaluate models with accuracy, recall, and F1-score, and tune hyperparameters to fix overfitting and underfitting.
• Capstone: turn a business problem into an end-to-end data science solution, from messy data to a working predictive model.