I have a variable called 'list_price' in my data ,which is not normally distributed.

Here is the screenshot of how the distribution looks like before transformation

enter image description here

Tried transforming it with the following transformations: 1.log 2.Sqrt 3.reciprocal 4.boxcox

None has resulted in normal distribution.Most of the transformations gave a skewed distribution after transformation. Would like to use this variable in modelling.

How would we proceed further in such scenarios? Should we go with original distribution itself?

Need some guidance on this

  • $\begingroup$ Why do you want a normal distribution? The most common time to want a normal distribution in modeling is to have a normal distribution of the error term of a regression, not a normal distribution of the data themselves. $\endgroup$
    – Dave
    Aug 10 '20 at 0:46

Based on your image, your data looks like it is uniformly distributed. You might try using more than 3 bins to look at the distribution more closely.

A few points related to your questions:

  1. For uniformly distributed data, you won't be able to transform this data into a valid normal distribution
  2. This distribution is fine for use in models, no need to change it.
  3. Here is a good resource explaining assumptions about linear regression, this information also applies to many other algorithms

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.