Datasets:

DOI:
License:
DS-5 / README.md
jcctello's picture
Update README.md
df81002 verified
---
license: gpl-3.0
---
# Simulated Light Curves for Strong Gravitational Lensing
This mock dataset simulates light curves for strong gravitational lensing (optical data). Each file contains pairs of light curves with a time delay of five days (**DS-5**).
### Table: Simulated Large Scale Data Sets
| Noise | Gap Size 0 | Gap Size 1 | Gap Size 2 | Gap Size 3 | Gap Size 4 | Gap Size 5 |
|---------|-----------|-----------|-----------|-----------|-----------|-----------|
| 0% | 1 | 10 | 10 | 10 | 10 | 10 |
| 0.036% | 50 | 50 | 50 | 50 | 50 | 50 |
| 0.106% | 50 | 50 | 50 | 50 | 50 | 50 |
| 0.466% | 50 | 50 | 50 | 50 | 50 | 50 |
| **Sub-Total** | **151** | **1510** | **1510** | **1510** | **1510** | **1510** |
Total = **7,701 data sets per underlying function.**
5 underlying functions yield **38,505 data sets.**
## File Naming Convention
Files follow this notation:
DS-5-`<Function>`-GAP-`<GapSize>`-`<Realization>`-N-`<NoiseLevel>`
### Explanation of Components:
- **DS-5**: Indicates that the true time delay is five days.
- **Function (`<Function>`)**: Represents the underlying function number (ranges from `1` to `10`).
- **GAP (`<GapSize>`)**: Indicates the number of removed points:
- `0`: No gap.
- `1-5`: Different gap sizes.
- **Realization (`<Realization>`)**: Represents different random gap realizations (ranges from `1` to `10`).
- **N (`<NoiseLevel>`)**: Represents the noise level:
- `0`: No noise.
- `1-3`: Different noise levels.
### Example
DS-5-1-GAP-0-1-N-0
- `DS-5`: True time delay of five days.
- `1`: First underlying function.
- `GAP-0`: No gaps.
- `1`: First realization.
- `N-0`: No noise.
Each pair of light curves is divided into five blocks before applying the gap procedure.
---
### Plots
Noise level 2 = 0.106%
<img src="https://cdn-uploads.huggingface.co/production/uploads/679190b1013258a9da0965db/xNQ8qjs4XIgwn-ECLC2uZ.jpeg" alt="Noise level 2 = 0.106%" style="width: 35%;">
Noise level 3 = 0.466%
<img src="https://cdn-uploads.huggingface.co/production/uploads/679190b1013258a9da0965db/kRZDleX3yQn9yC7iJRQar.jpeg" alt="Noise level 3 = 0.466%" style="width: 35%;">
Underlying fuction 1
<img src="https://cdn-uploads.huggingface.co/production/uploads/679190b1013258a9da0965db/CjfAI6ngJk4rpyvQtYAHj.jpeg" alt="Underlying fuction 1" style="width: 35%;">
Underlying fuction 2
<img src="https://cdn-uploads.huggingface.co/production/uploads/679190b1013258a9da0965db/8PiXpAkBp6-lNXMBBndgK.jpeg" alt="Underlying fuction 2" style="width: 35%;">
Underlying fuction 3
<img src="https://cdn-uploads.huggingface.co/production/uploads/679190b1013258a9da0965db/9KUf8kEg9QtBHNVfVrAQT.jpeg" alt="Underlying fuction 3" style="width: 35%;">
Underlying fuction 4
<img src="https://cdn-uploads.huggingface.co/production/uploads/679190b1013258a9da0965db/gF8vcdedAsYBTUkRFV6fY.jpeg" alt="Underlying fuction 4" style="width: 35%;">
Underlying fuction 5
<img src="https://cdn-uploads.huggingface.co/production/uploads/679190b1013258a9da0965db/8eW8YpmHEY8_FSGWwE-Zl.jpeg" alt="Underlying fuction 5" style="width: 35%;">
Underlying fuction 6
<img src="https://cdn-uploads.huggingface.co/production/uploads/679190b1013258a9da0965db/KHM3iZq1SNA1oC4jYpI-K.jpeg" alt="Underlying fuction 6" style="width: 35%;">
Underlying fuction 7
<img src="https://cdn-uploads.huggingface.co/production/uploads/679190b1013258a9da0965db/l1qFOC7dh2GE8TxQiC9lD.jpeg" alt="Underlying fuction 7" style="width: 35%;">
Underlying fuction 8
<img src="https://cdn-uploads.huggingface.co/production/uploads/679190b1013258a9da0965db/mlpMSAjqcY5R2kV_-zboG.jpeg" alt="Underlying fuction 8" style="width: 35%;">
Underlying fuction 9
<img src="https://cdn-uploads.huggingface.co/production/uploads/679190b1013258a9da0965db/mFX4B_DBOmtVZlWjlQZQF.jpeg" alt="Underlying fuction 9" style="width: 35%;">
Underlying fuction 10
<img src="https://cdn-uploads.huggingface.co/production/uploads/679190b1013258a9da0965db/u_cctWxu7I-zOH0IJ7Oso.jpeg" alt="Underlying fuction 10" style="width: 35%;">
If you use this dataset, please cite the following paper:
> **Cuevas-Tello, J. C., Tiňo, P., Raychaudhury, S., Yao, X., & Harva, M.**
> *Uncovering delayed patterns in noisy and irregularly sampled time series: An astronomy application.*
> Pattern Recognition, **43**(3), 1165-1179 (2010).
> DOI: [10.1016/j.patcog.2009.07.016](https://doi.org/10.1016/j.patcog.2009.07.016)
### BibTeX Citation
```bibtex
@article{CUEVASTELLO20101165,
title = {Uncovering delayed patterns in noisy and irregularly sampled time series: An astronomy application},
author = {Juan C. Cuevas-Tello and Peter Tiňo and Somak Raychaudhury and Xin Yao and Markus Harva},
journal = {Pattern Recognition},
volume = {43},
number = {3},
pages = {1165-1179},
year = {2010},
issn = {0031-3203},
doi = {10.1016/j.patcog.2009.07.016}
}