# predict()

*noun*

say it: **“prih-DIKT”** **/prɪˈdɪkt/**

In scikit-learn, the method that uses a trained model to guess the answer for new examples, one guess per row.

## What does **predict()** do in scikit-learn?

**predict()** uses a fitted model to guess an answer for each row of new features: `model.predict(X)` hands back one
guess per row. The model must have been trained with `fit()` first.

```python
from sklearn.linear_model import LinearRegression
model = LinearRegression().fit([[60], [70], [80]], [20, 40, 60])
print(model.predict([[75], [90]]))   # [50. 80.]: cups for a 75° and a 90° day
```

## In a sentence

- “She passed next week's forecast to **predict()** to plan how much lemonade to make.”
- “**predict()** guessed 50 cups for a 75-degree day.”

Level: Apprentice · [More Python terms at this level](https://www.ask-codey.com/glossary?category=python&level=2)

Updated: November 19, 2026

Source: https://www.ask-codey.com/glossary/predict
