Python中.py文件和.ipynb文件的区别详解

# 查看.ipynb文件结构
$ jupyter nbconvert --to script example.ipynb
$ file example.ipynb  # 显示: JSON text data

# 学生成绩分析

```python
import pandas as pd
df = pd.read_csv('grades.csv')
df.describe()
```

![外链图片转存失败,源站可能有防盗链机制,建议将图片保存下来直接上传](https://img-home.csdnimg.cn/images/20230724024159.png?origin_url=output.png&pos_id=img-atMoNhZ5-1744465838720)

# grades_analysis.py
import pandas as pd
import matplotlib.pyplot as plt

df = pd.read_csv('grades.csv')
print(df.describe())

df.hist()
plt.savefig('output.png')
print("图表已保存至output.png")

# 单元1：数据加载
from sklearn.datasets import load_iris
iris = load_iris()

# 单元2：查看特征
print(iris.feature_names)  # 输出：['sepal length (cm)', ...]

# 单元3：训练模型
from sklearn.svm import SVC
clf = SVC().fit(iris.data, iris.target)

# debug_script.py
from sklearn.datasets import load_iris
from sklearn.svm import SVC

def main():
    iris = load_iris()
    print("Features:", iris.feature_names)  # 需手动添加调试输出
    
    clf = SVC().fit(iris.data, iris.target)
    # 需要添加更多print语句定位问题

if __name__ == "__main__":
    main()

# .ipynb转.py
$ jupyter nbconvert --to script notebook.ipynb

# .py转.ipynb
$ jupytext --to notebook script.py

project/
├── notebooks/        # 存放.ipynb文件
│   └── exploration.ipynb
├── src/              # 存放.py文件
│   └── utils.py
└── requirements.txt

# 在.ipynb中调用.py模块
%load_ext autoreload
%autoreload 2

from src.utils import data_cleaner

# 交互式使用模块功能
cleaner = data_cleaner()
cleaner.preview()  # 即时查看效果

$ jupyter nbconvert --execute --to notebook --inplace production.ipynb

# 安装过滤插件
$ pip install nbstripout

# 设置git过滤
$ nbstripout --install --attributes .gitattributes