+- The Lumin Archive (https://theluminarchive.co.uk)
+-- Forum: The Lumin Archive — Core Forums (https://theluminarchive.co.uk/forumdisplay.php?fid=3)
+--- Forum: Computer Science (https://theluminarchive.co.uk/forumdisplay.php?fid=8)
+---- Forum: Data Science & Simulation (https://theluminarchive.co.uk/forumdisplay.php?fid=27)
+---- Thread: Data Science & Simulation — Analysis, Modelling, Visualisation & Real-World Systems (/showthread.php?tid=100)
Data Science & Simulation — Analysis, Modelling, Visualisation & Real-World Systems - Leejohnston - 11-13-2025
Data Science & Simulation — Analysis, Modelling, Visualisation & Real-World Systems
Data Science combines mathematics, computing, and statistics to extract meaningful insights from data.
Simulation models recreate real-world systems so we can study behaviour, make predictions, and test scenarios.
This thread introduces the fundamentals in a practical, beginner-friendly way.
-----------------------------------------------------------------------
1. What Is Data Science?
Data Science involves:
• collecting data
• cleaning and organising data
• analysing trends
• building models
• visualising results
• making predictions
Example applications:
• predicting weather
• analysing stock markets
• detecting fraud
• modelling disease spread
• evaluating scientific experiments
-----------------------------------------------------------------------
2. Types of Data
Numerical data: numbers
• continuous (temperature)
• discrete (counts)
Categorical data: colours, labels, categories
Time series data: data measured over time
Text data: natural language, documents
Understanding data type → choose the right analysis method.
-----------------------------------------------------------------------
3. Essential Tools of Data Science
Popular tools:
• Python
• NumPy (numerical computing)
• pandas (data tables)
• Matplotlib / Seaborn (visualisation)
• SciPy (scientific computing)
• scikit-learn (machine learning)
• Jupyter notebooks
Python is the most widely used language for data science.
-----------------------------------------------------------------------
4. The Data Science Workflow
1. Collect data
• sensors, simulations, databases, surveys
2. Clean data
• remove errors
• fill missing values
• normalise formats
3. Analyse
• summary statistics
• correlations
• distributions
4. Model
• regression
• classification
• clustering
• forecasting
5. Visualise
• graphs
• heatmaps
• scatter plots
• dashboards
6. Interpret
• conclusions
• insights
• recommendations
-----------------------------------------------------------------------
5. Basic Statistical Concepts
• mean, median, mode
• range, variance, standard deviation
• correlation
• probability distributions
• confidence intervals
Understanding these supports every analysis.
-----------------------------------------------------------------------
6. Regression — Predicting Continuous Values
Regression finds the best-fit relationship between variables.
Example:
Predicting house prices from:
• size
• number of rooms
• location
The model learns:
Code:
price = a(size) + b(rooms) + c-----------------------------------------------------------------------
7. Classification — Predicting Categories
Used for:
• spam vs non-spam
• disease vs healthy
• yes/no decisions
• image recognition
Common models:
• logistic regression
• decision trees
• random forests
• support vector machines
-----------------------------------------------------------------------
8. Data Visualisation
Visuals help reveal patterns.
Useful graphs:
• line charts (trends over time)
• bar charts (comparisons)
• histograms (distribution)
• scatter plots (correlation)
• box plots (spread & outliers)
Good visualisation = clear structure + correct axes + readable labels.
-----------------------------------------------------------------------
9. Simulation — Modelling Real Systems
Simulations recreate complex behaviour so we can test ideas safely.
Examples:
• orbital mechanics
• population dynamics
• chemical reactions
• traffic flow models
• weather and climate systems
• disease spread (SIR models)
• physics engines in games
Simulations help explore “what if?” scenarios.
-----------------------------------------------------------------------
10. Types of Simulation Models
Deterministic: same input → same output every time
(e.g., classical physics)
Stochastic: includes randomness
(e.g., Monte Carlo simulations)
Agent-based models: individual agents follow rules
(e.g., modelling animals, people, robots)
Differential equation models: describe continuous systems
(e.g., population growth, disease spread)
-----------------------------------------------------------------------
11. Monte Carlo Simulation
Monte Carlo methods use random sampling to estimate results.
Used in:
• finance
• physics
• nuclear engineering
• AI
• risk assessment
Example:
Estimating π by random points inside a circle.
-----------------------------------------------------------------------
12. Common Mistakes in Data Science
❌ Not cleaning the data
✔ poor data = poor model
❌ Confusing correlation with causation
✔ correlation ≠ cause
❌ Overfitting the model
✔ model memorises instead of generalising
❌ Using incorrect visualisations
✔ wrong graph = misleading
❌ No test/train split
✔ leads to inflated accuracy
-----------------------------------------------------------------------
13. Practice Questions
1. Name two differences between regression and classification.
2. What is a Monte Carlo simulation used for?
3. Why is data cleaning important?
4. Explain one advantage of using Python for data science.
5. Give an example of deterministic vs stochastic simulation.
-----------------------------------------------------------------------
Summary
This post covered:
• data types
• tools
• workflow
• regression
• classification
• visualisation
• simulation theory
• Monte Carlo
• common mistakes
• practice questions
Data Science & Simulation form the bridge between raw information and deep understanding — essential for research, engineering, AI, and scientific discovery.