What's the Best Way to Go About Normalizing Data for Better Analysis? or How Important Is Normalizing

18 Replies, 1137 Views

Ugh, I feel you. Normalizing data is one of those things that seems simple until you’re deep in it.

Min-max scaling is great if you want everything between 0 and 1, but z-scores are better if outliers are wrecking your life.

Check out pandas for quick normalization—df[‘col’] = (df[‘col’] - df[‘col’].mean()) / df[‘col’].std()

Also, this guide helped me a ton: [link to a normalization tutorial].

Messages In This Thread



Users browsing this thread: 1 Guest(s)