Python Pandas - 检测列和数字格式

Question

Im using Pandas to manage csv. Unfortunally I have columns with number that use "comma" as decimal separator:

Es. 50,12

When I use convert_dtypes() function, this columns are converted to String and not number so the sort functions doesn't work properly.

Is there a way to specify "number format" of dataset so every number is considered like NNNN,DD instead of NNNN.DD

EXAMPLE:

| Gross Amount | Item Number|
-----------------------------
|52,50         |   1       |
|498,00        |   2       |
|10,01         |   3       |
|1,74          |   4       |
|518,04        |   5       |
|2,10          |   6       |

AutoDetect return this:

Gross Amount     string
Item Number       Int64

So When I order by "Gross Amount" it sort String and not number so, for example, "10,01" is printed before "2,10"

Answer 1

When suppose the DataFrame is called df . you can use the code below.

Code:

for column in df.columns:
    if pd.api.types.is_string_dtype(df[column]):
        try:
            df[column] = df[column].str.replace(',', '.').astype('float')
        except:
            pass