filename stringclasses 8
values | rows int64 500 4k | columns int64 500 4k | dtype stringclasses 1
value | seed int64 42 42 | size_MB float64 1.91 122 |
|---|---|---|---|---|---|
matrix_500.bin | 500 | 500 | float64 | 42 | 1.91 |
matrix_1000.bin | 1,000 | 1,000 | float64 | 42 | 7.63 |
matrix_1500.bin | 1,500 | 1,500 | float64 | 42 | 17.17 |
matrix_2000.bin | 2,000 | 2,000 | float64 | 42 | 30.52 |
matrix_2500.bin | 2,500 | 2,500 | float64 | 42 | 47.68 |
matrix_3000.bin | 3,000 | 3,000 | float64 | 42 | 68.66 |
matrix_3500.bin | 3,500 | 3,500 | float64 | 42 | 93.46 |
matrix_4000.bin | 4,000 | 4,000 | float64 | 42 | 122.07 |
Generated Matrices for SVD Experiments
Description
This dataset contains generated numerical matrices for Singular Value Decomposition (SVD) experiments.
Data Format
- Format: Raw binary (.bin)
- Data type: float64
- Storage order: Row-major (C-style)
- Dimensions: See filenames and metadata.csv
- Origin: Synthetic generated data
How to Load a Matrix
Example for a 500 x 500 matrix:
import numpy as np
A = np.fromfile("matrix_500.bin", dtype=np.float64)
A = A.reshape((500, 500), order="C")
print(A.shape)
Use the appropriate dimensions for each matrix file.
Intended Use
- Singular Value Decomposition (SVD)
- Matrix reconstruction
- Numerical linear algebra
- Performance experiments
Limitations
These matrices are synthetic and do not represent real-world measurements. Verify each matrix's dimensions and binary format before loading it.
- Downloads last month
- 16