Review the detected table before normalizing values
Use this tool to standardize selected missing-value markers into one target representation. It does not infer missing data or write replacement values. Table processing and export occur in the browser.
Configure markers, targets, and column rules
Choose whether Empty string and Whitespace-only cells count as missing, then select built-in markers such as NULL, N/A, NA, unknown, - (dash), and none. Add other values in Custom missing-value markers (comma-separated). Matching is case-insensitive.
Select Empty string, NULL, N/A, NA, - (dash), or Custom... as the target. After the columns are detected automatically, enable an override for any column that needs a different target. Use First row is a header to control whether the first row supplies column names.
How to normalize null values
- Paste a table into Paste your table below, or drop/upload a CSV, TSV, or TXT file. If the input is empty, the tool shows an alert.
- Select the missing-value markers, add custom markers if needed, and choose a target representation.
- Columns are detected automatically after data is pasted or uploaded.
- Review the detected columns and set any per-column overrides.
- Click Normalize Null Values.
- Review the counts, preview, and change log, then use Export CSV for the normalized table or the change-log export for its audit file.
Review the preview and change log
The on-page preview shows the first 200 rows of the normalized result. The visible change log shows at most 500 changes, although its count can represent the full result. The normalized CSV export includes all processed rows, and the exported change-log CSV includes the full change list.
Understand parsing and export limitations
The tool accepts CSV, TSV, and TXT input and detects comma, tab, semicolon, or pipe delimiters. The normalized export is always a comma-separated CSV, regardless of the input delimiter.
When rows have different lengths, detection uses the most common column count: short rows are padded and excess fields are merged. The output is therefore not a byte-for-byte copy of the input. Export always adds a header row; when First row is a header is unchecked, it uses generated names such as Column 1 and Column 2. These parsing and export rules can affect table structure.
Browser-side processing and practical scope
Table processing and file export occur in the browser. This page does not establish account requirements or usage limits. The tool standardizes markers; it does not impute replacement data. The related Fill Blank Cells tool is intended for writing replacement values, but its available rules are not established here.
Frequently Asked Questions
How is this different from Fill Blank Cells?
This tool only standardizes how missing-value markers are written - for example turning "N/A", "null", and "-" into one consistent marker - without inventing any new data. The related Fill Blank Cells tool is intended for writing replacement values into blank cells; see that tool for its available rules.
What counts as a missing-value marker?
You control this directly: toggle whether truly empty strings and whitespace-only cells count, check off common markers like NULL, N/A, NA, unknown, dash, or none, and optionally add your own custom markers. Matching is case-insensitive. Only cells matching your definition are normalized.
Can different columns use different representations?
Yes. After the columns are detected automatically, you can override the target representation for any individual column - for example normalizing missing values to an empty string in most columns but to "N/A" in one specific column - while all other columns use the global default you set.
Can I see exactly what changed before exporting?
The preview highlights normalized cells in its first 200 rows. The on-page change log displays up to 500 changes and shows the row, column, original marker, and new value. The exported change-log CSV contains the full change list.
Is the normalized export always CSV?
Yes. The normalized export is always a comma-separated CSV, even when the input uses tabs, semicolons, or pipes.