1. 数据准备

```python import pandas as pd df = pd.read_csv('datasets/dpc-covid19-ita-regioni.csv') df.dropna(axis=1, inplace=True) df.tail(10) ```

2. Z 得分

```python from scipy.stats import zscore df['zscore_deceduti'] = zscore(df['deceduti']) df['zscore_deceduti'] ```

3. Z 映射

```python from scipy.stats import zmap df['zmap_deceduti'] = zmap(df['deceduti'], df['terapia_intensiva']) df['zmap_deceduti'] ```

4. 异常值检测

```python import numpy as np threshold = 2 df['outliers'] = np.where((df['zscore_deceduti'] - threshold > 0), True, np.where(df['zscore_deceduti'] + threshold < 0, True, False)) df['outliers'] ```

5. 结论

在本教程中，我们了解了规范化和标准化之间的区别。标准化数据可以通过两种方法：Z 得分和 Z 映射。标准化可用于检测和删除数据集中的异常值，也可以用于在不同的数据集之间进行比较。

---

python pandas 怎样高效地添加一行数据

定义行 row = (row_values,column_headers)追加行 (row,ignore_index=True)举例 :row = ([python, pandas, 10], [name,lib,rank])(row,ignore_index=True)

python3.6怎么单独正则化/标准化DataFrame中的指定列数据

dataframe是神马？我只会json格式的数据转dict类型123importjsonj=(\\{name:123,Age:22}\\)print(j)========你是指pandas里的dataframe吗？传入dict参数实例化一下就行了