如何基于含缺失值的Pandas DataFrame创建新数据表？

Hey there! Looking at your sample DataFrame with those NaN values, I’m guessing you want to create a new DataFrame by handling those missing entries in some way. Let’s walk through a few common use cases that might match what you’re after:

If you want to replace NaNs with specific values or calculated metrics, here are a few practical options:

Fill with a constant value (e.g., 0) #

df_filled_zero = df.fillna(0)

Result:

Col_A  Col_B  Col_C

2 0.001 0.00 0.4566
3 0.250 0.46 0.0000
4 1.000 0.00 2.0000
5 0.000 0.00 5.5000

Fill with column means (for numeric columns) #

df_filled_mean = df.fillna(df.mean(numeric_only=True))

Result:

Col_A  Col_B  Col_C

2 0.001 0.46 0.4566
3 0.250 0.46 2.6589
4 1.000 0.46 2.0000
5 0.417 0.46 5.5000

Forward-fill (use the previous row's value) #

df_filled_ffill = df.fillna(method='ffill')

Result:

Col_A  Col_B  Col_C

2 0.001 NaN 0.4566
3 0.250 0.46 0.4566
4 1.000 0.46 2.0000
5 1.000 0.46 5.5000

If you want to filter out rows or columns that have any NaNs:

Drop rows with any missing values #

df_clean_rows = df.dropna(axis=0)

Note: Since every row in your sample has at least one NaN, this will return an empty DataFrame.

Drop columns with any missing values #

df_clean_cols = df.dropna(axis=1)

Note: Similarly, every column in your sample has NaNs, so this will also return an empty DataFrame.

If you want a boolean DataFrame that marks where NaNs are located:

df_missing_flags = df.isna()

Result:

Col_A Col_B Col_C
2 False True False
3 False False True
4 False True False
5 True True False

If you had a specific transformation in mind (like calculating a new column based on existing ones, or reshaping the data), feel free to share more details!

内容的提问来源于stack exchange，提问作者Kdog