Reconstruct. This could only be employed if you have an individual dtype for all columns. If not, you may end up with dtype object for all columns and converting them again involves more dictionary function. df.columns = new in which new is the list of new columns names is as https://raymondbbawo.bloggerchest.com/36533780/not-known-details-about-stack-ball-levels-challenge