Data Tools tutorials and worked examples

Choose charts from the question, data and comparison.

• Choose a chart for a dataset by naming the comparison

  Create category, time-series and paired-observation charts from explicit data, and check how aggregation can hide the story.

  tutorial · 4 min read

Understand shape, memory and numerical operations with inspectable arrays.

• NumPy broadcasting rules: align axes before adding arrays

  Predict broadcast shapes from the right, normalize columns correctly and diagnose an accidental outer operation with checked arrays.

  tutorial · 4 min read

• NumPy views vs copies: predict which array changes

  Trace shared memory through slicing, advanced indexing and assignment, and set an explicit copy boundary in a numerical workflow.

  troubleshooting · 4 min read

• NumPy projects for beginners: three array projects with complete code

  Build three NumPy projects with complete Python code: summarise weekly sales, clean sensor readings and scale features. Understand shapes, masks and broadcasting.

  project · 9 min read

Clean, combine and inspect tables while preserving their meaning.

• pandas groupby aggregation vs transform: group rows or preserve them

  Compute group totals and row-level shares from an explicit table, including missing keys and a zero-total group.

  tutorial · 4 min read

• pandas loc vs iloc: labels, positions and safe assignment

  Use a nonconsecutive integer index to expose label versus position selection, slicing endpoints and aligned assignment.

  tutorial · 4 min read

• Why a pandas merge duplicates rows: validate the keys first

  Calculate a join's row count, reproduce many-to-many expansion, and use validate and indicator to enforce the intended table relationship.

  troubleshooting · 4 min read

• Three pandas projects for beginners, with complete code and output

  Build three pandas projects with complete Python code: clean orders, join product tables and calculate a daily trend. Check outputs and common data mistakes.

  project · 9 min read

Connect preparation, fitting and evaluation without crossing data boundaries.

• scikit-learn fit, transform and predict: keep learned state inside training

  Run an imputer-scaler-classifier pipeline, inspect fitted statistics and verify that prediction does not refit preprocessing on held-out data.

  tutorial · 4 min read

• Machine learning code in Python: your first model in about 30 lines

  A complete, runnable scikit-learn example: load a built-in dataset, split it, beat a baseline, fit a model and evaluate it, with every line explained.

  tutorial · 8 min read