EDA Label Standardization - Time Series Mechanics

Read the Cleanup

Start with five pinned rows. Email and email should mean the same category, and Ads and ads should also mean the same category. Normalize the labels to lowercase. After cleaning, the categories count as email = 3 and ads = 2.

A tiny label-cleaning readout: start with mixed capitalization, standardize category labels, then compare category counts before and after cleaning.

highlighted = computed this step

Load mixed labels

Start with 5 pinned rows where label capitalization is inconsistent.

\text{mixed labels}

Count before cleaning

Before cleaning, Email and email count as different labels; Ads and ads do too.

\text{before counts}

Standardize labels

Normalize each label to lowercase so Email and email both become email.

\text{lowercase labels}

Count after cleaning

After cleaning, the count readout is email = 3 and ads = 2.

\text{after counts}