One Hot Encoding with multiple columns

How can I update a dataset that was already been hot encoded once? For example, there are two columns A and B both with similar data that needs to be hot encoded. I understand using the One to Many Node to hot encode and execute hot encoding on Column A. But how can I then take the data that is in Column B to update the newly formed hot encoded columns?

In other words, what is a good way to perform hot encoding when there are multiples values to be encoded and not just 1? Thanks!

Do you still have all of the original columns?

If so, you can just one-hot encode all the columns

For example, if I start with this table:
image

I can one-hot encode every column to get this table:

You’ll notice that we’ve just added extra columns for each of the original columns we want to encode.

If you don’t have the initial data, you can just encode the column you do have, append these new columns to the already encoded table, and modify the names so that the source column of each set of encoded values is obvious.

2 Likes

@rinaldiinjapan in case you are looking for other encoding methods this might give you further ideas

2 Likes

Love the simplicity. Thanks!

2 Likes