DocsData CleaningDynamic Interactive Cleaning

CUSTOM RULESData Cleaning Engine

Dynamic Interactive Cleaning

Choose custom rules per column, preview the live before-and-after diff, and apply changes with confidence.

Live Before-and-After Diff Preview

Never guess what a rule will do. DataForge shows you an interactive side-by-side diff comparing your original data against the clean version before you commit any changes.

What is this?

Dynamic Interactive Cleaning gives you complete, surgical control over your data. Rather than applying a blanket rule to everything, you choose the exact column you want to fix and pick the rule that fits your business logic. Want to replace missing discounts with 0, change customer names to UPPERCASE, mask phone numbers for privacy, or filter out negative prices? Dynamic Cleaning lets you set the rule, inspect a live before-and-after preview, and apply it with confidence.

Step-by-Step Walkthrough

Select Your Column

Pick one or more columns from the list. You can use the search bar or type filters (numeric, text, datetime, missing) to find columns quickly.

Choose an Action Category

Select from Impute (fill blanks with median, mean, mode, or custom value), Outlier (clip or drop extreme spikes), Scale, Encode, Text/PII (trim, casing, regex, phone/email masking), or Type Cast.

Click 'Preview Diff'

DataForge runs a fast preview comparing your current values with the proposed changes, showing exact numbers and row counts before anything touches your dataset.

Click 'Apply Rule'

When you're happy with the preview, click 'Apply Rule'. DataForge applies the change, adds it to your recipe history, and updates your table immediately.

Why Use It

Zero guesswork: you see exactly what will happen to your data before committing any change.

Full control: customize cleaning rules to match your specific business requirements.

Unlimited undo: step backward through your rule history anytime if you change your mind.

Where It Can Be Used

Retail Catalog Management

Product SKU & Price Normalization

Force SKU strings to uppercase, strip currency symbols, and ensure discount percentages stay between 0% and 100%.

Human Resources

Employee Compensation Formatting

Standardize department names ('sales' vs 'Sales') and impute missing bonus figures with a custom zero value.

Pro Analyst Tips from Our Team

Always click 'Preview Diff' before using rules that drop rows so you don't accidentally remove valuable customer records.

Use the PII Masking option under Text to automatically hide phone numbers, emails, and credit cards before sharing datasets with your team.

Want to test DataForge with your dataset first?

Start on the Free Starter plan — zero setup, instant automated EDA.

Explore Architecture