Whenever you're ready, walk me through how you'd approach this.
How would you handle missing data or outliers in a large dataset using Python pandas?
Explain how you would profile missingness and extreme values, decide whether to remove, impute, cap, transform, or retain observations, and validate that preprocessing does not distort the data. Show concise pandas code for a numeric and categorical dataset. Discuss how the approach changes when missingness is informative, outliers are legitimate rare events, or the data will be used for model training. Do not assume that a single global rule, such as dropping every incomplete row or removing values beyond a fixed standard-deviation threshold, is appropriate.