p parqview

Open-source terminal data inspection

Inspect Parquet and CSV files without leaving the terminal.

A keyboard-first viewer for browsing tabular data, checking columns, and spotting missing values before they cause downstream surprises.

Install parqview View on GitHub
  • Parquet and CSV
  • Keyboard-first
  • Built with Go and DuckDB
  • MIT licensed
A parqview session showing a dataset table alongside a searchable column list.

Keyboard-first workflow

Move through rows, columns, and panes without turning a quick check into a long session.

Column triage

Filter by name, build a selection set, and focus on the fields that matter for a given check.

Missing-value checks

Surface nulls and NaNs visually and step through them while validating a file.

Easy handoff to Python

Copy selected columns as a Python list to carry straight into a notebook or script.

Column summaries

View descriptive stats, top values, and distributions for a quick sense of any column's shape.

Before training a model

Confirm expected columns are present and check missing values.

After a join or merge

Scan for empty columns, duplicated fields, or key mismatches.

While narrowing a dataset

Search and select columns in large files.

Install

Homebrew

Recommended on macOS.

brew install robince/tap/parqview

Run

Open a Parquet or CSV file from the terminal.

parqview path/to/file.parquet