# GDS Exercise 3 – 2025 — Data

来源：https://www.kaggle.com/competitions/gds-exercise-3-2025-/data

本地 Meta Kaggle 快照；不是本次实时页面。

# Dataset Description

## Files

- **train.csv** – your labeled set of data, consisting of our *existing* knowledge of habitable exoplanets.

```python
import pandas as pd

data = pd.read_csv('train.csv')
X_train = data.drop(columns=['y_train'])  # DataFrame of shape (n_samples, 372) where each row represents an exoplanet and each column is a feature.
y_train = data['y_train']                # Series of shape (n_samples,), containing binary habitability labels (1 for potentially habitable, 0 otherwise).
```

- **test.csv** – exoplanets in the test set, but with unknown habitability (without labels)—it is your job to predict these.

- **sample_submission.csv** – a sample submission file in the correct format for reference.

When preparing your submission, load sample_submission.csv and insert your predictions to validate the order of the "ID" column. For example:

```python
sample = pd.read_csv('sample_submission.csv')
sample['P_HABITABLE'] = y_pred  # assuming y_pred contains your predictions
sample.to_csv('submission.csv', index=False)
```

- **metaData.csv** – supplemental information about the data

- `y_train` – Binary label indicating exoplanet habitability:
  - `1`: The exoplanet is potentially habitable.
  - `0`: The exoplanet is non-habitable.

- **exampleESI.csv** – an example file consisting of 50 exoplanets with their `ESI` values for reference.