Skip to content

CSV column select

Pick the columns you want from a CSV or TSV, in the order you want them, and optionally rename them.

CSVcolumnsrename

Upload

Files (0–1)
One delimited file, up to 8 MB, or paste it below.
Paste a delimited document to extract from instead.

Options

Columns to extract Comma-separated, in the order you want them. The order you list the columns is the output order; a name may be listed twice if the renamed outputs stay distinct.
Rename the output columns Optional. One new name per selected column, same order; renaming requires a headered document and unique, non-empty names.
Questions people ask
Can I reorder columns too?
Yes: the order you list the columns is the order the output carries them ("name, email" extracts name first), whether you keep all columns or just a few.
How do renames work?
Give one new header name per selected column, in the same order: selection "c,a" with rename "gamma,alpha" produces a table headed gamma, alpha. Renaming is all-or-nothing for the selection, needs a headered document, and the final names must be non-empty and unique; otherwise the run refuses with a named diagnostic.
My file has a column twice. What happens?
A column NAME that matches two or more columns is ambiguous and refuses the run (naming the positions). It is never a silent first match. Rename the duplicates first by selecting them and giving distinct output names.
Is my data retyped or cleaned?
Never. A cell's value is exactly the parsed field: "007" stays "007", "true" stays "true", leading and trailing spaces stay. Only the column layout changes.
What happens when a row is ragged?
With the default strict policy the run is refused and one diagnostic names every disagreeing row. Permissive extracts anyway: a position the row does not have emits an empty cell, and every fill is counted in the result.