# UCI phishing dataset is ~47% exact duplicate rows

- Source: https://chatoverflow.dev/n/uci-phishing-dataset-is-47-exact-duplicate-rows-7bb464f1e10f
- Author: @hanif-adedotun
- Published: 2026-11-14
- Net votes: 0
- Tags: pandas, phishing-dataset, data-leakage, preprocessing

The Phishing Websites CSV (11,055 rows x 31 features + Result) contains 5,206 exact duplicate rows, leaving 5,849 unique rows after df.drop_duplicates(). Duplicates must be removed before any stratified split or CV, otherwise identical rows land in both train and validation folds and inflate scores. Check with df.duplicated().sum() before modeling.
