Cheng-Ju Wu /
← Back to portfolio
R / DATA ANALYTICS / 2025

Making patterns useful.

Selected R-based work connecting analytical methods with interpretable business questions. All three projects were completed in 2025.

01 / R · K-means · Hierarchical clustering · 2025

Australian occupation clustering

The question

How do occupations differ when employment, earnings and education are considered together?

What I did

Integrated and standardised Australian occupation data, then applied K-means and hierarchical clustering. Visualised the resulting profiles using charts, heatmaps and box plots.

What it tells us

The analysis identified distinct multivariate occupation profiles. These are descriptive groups, not forecasts of growth or automation risk.

02 / R · Time series · Data visualisation · 2025

Electricity demand time series

The question

How does electricity demand change across the day, the week and unusual events?

What I did

Analysed half-hourly Australian electricity-demand data, visualised daily, weekly and monthly patterns, and compared a storm-event day with typical February demand profiles.

What it tells us

The work describes observed demand patterns and abnormal periods. It does not claim a forecasting model or isolate the causal effect of the storm.

03 / R · Market analysis · ROI · 2025

Movie market analytics

The question

How are genre, release timing and budget associated with box-office performance?

What I did

Integrated movie budget and box-office datasets to examine relationships between genre, release timing, budget, revenue and ROI.

What it tells us

Results are observed associations. Box-office-based ROI is a comparison measure and does not represent audited film profitability.

Additional Power BI dashboard experience is included in my skill set; the case studies here are R-based.