Suppose the following df:

     Airplane_name   col2      col3     ...     colm
[1]     2352365236
[2]     6235231555
[3]     3463463435
[n]     6497830120

Lets suppose that n=10^6.

What is the best way to work the airplane_name column?

One-hot encoding would cause too many columns. Should I leave that column as is, but convert it to a string?

Is it different for unique int values ​​than for unique string values? If the answer is "yes", what do you recommend?

  • 1
    $\begingroup$ Does this answer your question? Can identifiers be used to train a model? $\endgroup$
    – noe
    Mar 28 at 7:06
  • $\begingroup$ It gives me an intuitive idea of ​​how to solve the problem, thank you very much! $\endgroup$
    – Daniel_DS
    Mar 28 at 17:04
  • $\begingroup$ please, consider upvoting questions and answers that are useful to you. $\endgroup$
    – noe
    Mar 28 at 17:11


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.