How do I convert my object type data to numeric in pandas?

Question

I have a dataset like below. Some of the attributes are in object type data. I want to convert them to numeric tye data. but when I use pd.to numeric function, all the object type data converts to 'nan'

                                          Flow ID    Source IP  ...  Idle Min   Class
    523320       10.42.0.211-10.42.0.1-3145-53-17  10.42.0.211  ...         0  BENIGN
    502082       10.42.0.151-10.42.0.1-9436-53-17  10.42.0.151  ...         0  BENIGN
    579158  192.229.173.173-10.42.0.42-80-50469-6   10.42.0.42  ...         0  BENIGN
    59163       10.42.0.211-10.42.0.1-39702-53-17  10.42.0.211  ...         0  Malwre
    217478       10.42.0.151-10.42.0.1-4616-53-17  10.42.0.151  ...         0  Malwre

The data type is shown below

    Flow ID              object
    Source IP            object
    Source Port           int64
    Destination IP       object
    Destination Port      int64

    Idle Mean           float64
    Idle Std            float64
    Idle Max              int64
    Idle Min              int64
    Class                object

Now I want to use all my data and convert all the data into int or float.

@Carsten Actually I have no idea, I wanted to use this dataset for Machine Learning. So I wanted to convert this data. — Nasimul Hasan Chowdhury, Apr 07 '20 at 06:18
You could use sklearn's LabelEncoder for it, check out [this](https://stackoverflow.com/questions/34915813/convert-text-columns-into-numbers-in-sklearn) post for details — RishiC, Apr 07 '20 at 06:20
Without seeing your full dataset (all columns), it's not possible to advise on all data types. It appears as though pandas has it right (from what I can see). An IP address is certainly **not** and `int`, but rather a string, or `object`. — S3DEV, Apr 07 '20 at 06:27
@S3DEV you are right, but my question is how do I use this type data? — Nasimul Hasan Chowdhury, Apr 07 '20 at 06:34

How do I convert my object type data to numeric in pandas?

0 Answers0