How to detect and remove outliers within a dataset using boxplot?

Question

I want to detect and remove outliers from a stock prediction dataset.Whether I have to take one columns or whole dataset?