README.md

---
license: cc-by-4.0
pretty_name: "Kepler Transit Timing Catalog"
language:
  - en
description: "Holczer et al. (2016) Kepler transit timing catalog — 295,187 individual transit times for 2,599 Kepler Objects of Interest (KOIs), with O-C residuals, durations, and depths."
task_categories:
  - tabular-regression
tags:
  - space
  - exoplanets
  - kepler
  - transit-timing
  - ttv
  - astronomy
  - open-data
  - tabular-data
size_categories:
  - 100K<n<1M
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/kepler_transit_timing.parquet
    default: true
---

# Kepler Transit Timing Catalog

*Part of the [Astronomy Datasets](https://huggingface.co/collections/juliensimon/astronomy-datasets-69c24caf2f17e36128946743) collection on Hugging Face.*

Transit timing catalog from Holczer et al. (2016), containing **295,187** individual transit
mid-times for **2,599** Kepler Objects of Interest (KOIs). Each record includes the
observed mid-transit time, observed-minus-computed (O-C) residual, transit duration, and
transit depth with uncertainties.

## Dataset description

Transit timing variations (TTVs) occur when gravitational interactions between planets in a
multi-planet system cause measurable deviations from a strictly periodic transit schedule.
Holczer et al. (2016) performed a uniform analysis of all Kepler long-cadence light curves
to extract individual transit times, producing the most comprehensive Kepler TTV catalog.
The O-C (observed minus computed) residuals reveal planetary interactions, orbital
eccentricities, and the presence of additional non-transiting planets.

## Key columns

| Column | Type | Description |
|--------|------|-------------|
| `koi` | float64 | Kepler Object of Interest number |
| `transit_number` | Int32 | Sequential transit number for this KOI |
| `t_obs_bjd` | float64 | Observed mid-transit time (BJD - 2454833) |
| `t_obs_err` | float64 | Uncertainty on mid-transit time (days) |
| `o_c` | float64 | Observed minus computed residual (days) |
| `o_c_err` | float64 | Uncertainty on O-C residual (days) |
| `duration_hr` | float64 | Transit duration (hours) |
| `duration_err` | float64 | Uncertainty on transit duration (hours) |
| `depth_ppm` | float64 | Transit depth (ppm) |
| `depth_err` | float64 | Uncertainty on transit depth (ppm) |

## Quick stats

- **295,187** individual transit times
- **2,599** unique KOIs
- Median O-C residual: **0.0000** days
- Median transit depth: **nan** ppm
- Median transit duration: **nan** hours

## Usage

```python
from datasets import load_dataset

ds = load_dataset("juliensimon/kepler-transit-timing", split="train")
df = ds.to_pandas()

# TTVs for a specific KOI
koi_137 = df[df["koi"] == 137.01].sort_values("transit_number")
print(f"KOI 137.01: {len(koi_137)} transits")

# Plot O-C diagram
import matplotlib.pyplot as plt
plt.errorbar(koi_137["transit_number"], koi_137["o_c"],
             yerr=koi_137["o_c_err"], fmt=".", ms=3)
plt.xlabel("Transit number")
plt.ylabel("O-C (days)")
plt.title("KOI 137.01 Transit Timing Variations")
plt.show()

# KOIs with the strongest TTVs (largest O-C scatter)
ttv_rms = df.groupby("koi")["o_c"].std().sort_values(ascending=False)
print("Top 10 TTV candidates:")
print(ttv_rms.head(10))
```

## Data source

Holczer, T. et al. (2016), "Transit Timing Observations from Kepler. IX. Catalog of
Transit Timing Measurements of the Long-Cadence Data", ApJS, 225, 9. Accessed via
[VizieR](https://vizier.cds.unistra.fr/), CDS Strasbourg (J/ApJS/225/9).

## Pipeline

Source code: [juliensimon/space-datasets](https://github.com/juliensimon/space-datasets)

## Citation

```bibtex
@dataset{kepler_transit_timing,
  author = {Simon, Julien},
  title = {Kepler Transit Timing Catalog},
  year = {2026},
  publisher = {Hugging Face},
  url = {https://huggingface.co/datasets/juliensimon/kepler-transit-timing},
  note = {Based on Holczer et al. (2016) ApJS 225, 9, via VizieR CDS}
}
```

## License

[CC-BY-4.0](https://creativecommons.org/licenses/by/4.0/)