Missing Data Handler
See exactly which cells are missing, then fill or drop them with one click.
Drag & drop your file here
or click to browse
Handle Missing Values in Your Dataset
Missing data is everywhere: blank cells in CSVs, empty Excel/ODS columns, null values in JSON exports. It breaks formulas, skews statistics, and crashes visualizations. DataScrub shows you exactly where the gaps are and gives you four ways to fix them.
How to Fix Missing Data
- Upload your file — the health report appears automatically.
- Review the health report — see which columns and rows have missing values.
- Choose a fill strategy — pick the approach that fits your data best.
- Apply and download — get your cleaned dataset in seconds.
Fill Strategies Explained
- Default Value — fill all blanks with a single value like "N/A", "0", or "Unknown".
- Per-Column — set a different fill value for each column, so names get "Unknown" while amounts get "0".
- Forward Fill — copy the previous row value downward. Ideal for time series where the last known value carries forward.
- Drop Rows — remove any row that contains a missing value. Use when completeness matters more than sample size.
When to Use Each Strategy
Use Default Value for simple, uniform datasets where one fill covers everything. Per-Column is better when different columns need different treatment — for example, text columns filled with "Unknown" and numeric columns filled with "0". Forward Fill works best for time series and ordered data where values repeat until they change. Drop Rows is the right choice when you need a complete dataset and can afford to lose some observations.
Tips for Handling Missing Data
- Always profile your data first — understand the pattern of missingness before deciding how to handle it.
- Do not blindly fill — think about what the missingness means. Is it random, or does it signal something important?
- Forward fill is perfect for sensor data and financial time series where values persist until updated.
Frequently Asked Questions
What is forward fill?
Forward fill (also called LOCF — Last Observation Carried Forward) copies the last non-empty value in a column down to the next empty cell. It is commonly used in time series data where a value stays the same until a new reading is recorded.
Should I drop or fill missing values?
It depends on your data. Drop rows if you need a complete dataset and missing values are rare. Fill if missing values are frequent and dropping would lose too much data. Always consider what the missingness means in context before choosing.
Can I undo changes?
Yes. You can re-upload the original file at any time to start over. Each operation works on a fresh copy, so your source data is never modified in place.
What does the mini-map show?
The mini-map is a visual grid where each cell represents a value in your dataset. Green cells contain data, and amber cells are missing. It gives you a quick visual overview of where the gaps cluster — whether they are random or concentrated in specific rows or columns.