Random Vector Generator
Create reproducible lists of random vectors with selectable dimensions and distribution modes (uniform, unit, or Gaussian). Each vector is rounded to a specified precision.
Also known as: random points · vector sampler
seeded
Presets
Output
About this tool, tips & examples
What it does
The Random Vector Generator produces n-dimensional vectors (1 to 10 dimensions) in three modes: uniform components within a min/max range, unit vectors (length 1, uniformly distributed directions), or Gaussian components. Control decimal precision, generate up to 1,000 vectors per run, and reproduce any batch from its seed.
Common use cases
- Machine learning examples — feature vectors and embeddings-shaped data for pipeline smoke tests (a high-dimensional preset included).
- Geometry and physics testing — unit vectors as unbiased random directions for rays, normals, and velocities.
- Data visualization — 2D/3D point sets for scatter plots and projection demos (2D uniform and 3D presets).
- Numerical algorithm tests — dot products, normalization, and distance functions verified against reproducible input.
Settings
- Dimensions — 1 to 10.
- Mode — uniform (with Min/Max bounds), unit, or Gaussian.
- Decimals — 0 to 10 places.
- How many — 1 to 1,000 vectors, exportable as text, CSV, or JSON.
- Seed — identical seed + settings = identical vectors.
Privacy note
Vectors are computed locally in your browser; nothing is uploaded — pure synthetic mathematics.
FAQ
Why use unit mode for random directions? Naively normalizing uniform-cube samples biases directions toward the cube’s corners. Unit mode produces properly uniform directions on the sphere — the difference matters in simulations and graphics.
When Gaussian instead of uniform? Gaussian components make vector lengths concentrate predictably and are the standard model for noise and embeddings; uniform gives strictly bounded components. Match whatever your algorithm assumes.
More than 10 dimensions or nested shapes? The N-Dimensional Array generator does arbitrary tensor shapes; the Matrix generator covers 2-D grids with CSV export.