Is there any good way to deal with these abnormal data? Reject or replace. Thank you!
Solved! Go to Solution.
Hi @longhui ,
You can create a method to identify values that doesn't seem right or too high/low. (Outlier detection)
Luckily, @MarqueeCrew have developed an excelent set of tools.
http://www.chaosreignswithin.com/
The one you need is the modify outliers tool.
https://gallery.alteryx.com/#!app/CReW-Modify-Outliers/5a80861feffc2a2658b0ad96
Take a look 🙂
Best,
Fernando Vizcaino
Thank yon very much!