VISIDATA

With the right tool at his disposal, Shashank Sharma has no trouble making sense of vast heaps of tabular data.

Vast amounts of information, separated into logical rows and columns that add up to a V table, is the layman’s definition of a data set. We see such data sets everywhere, from the everchanging prices of consumer electronic goods to polling statistics, results of football matches, and more. With the right tool at your disposal, you can explore, clean, edit, summarise, analyse and plot the data into graphs for quick assimilation. While working with spreadsheets or CSV data is not traditionally envisioned as a task suited for the command line, VisiData will quickly dispel you of such notions.

Data from 2020 league results. You can use the mouse scroll wheel to navigate the file.

Written in Python and released under the GPLv3, VisiData is incredibly fast and lightweight. You can easily use it to work with data sets featuring a few million rows of data. The project’s goal for future releases is to be able to handle hundreds of millions of rows of data.

For our money, one of VisiData’s best features is that it’s non-destructive. This means that it doesn’t alter your raw data by default, leaving you free to manipulate the data to best analyse it without fear of changing it in any fashion. In fact, it’s also possible to save a VisiData session so that your particular manipulations are recorded, and you can then restore the session without having to start again from scratch. All this and more will become obvious as you go through this tutorial and, more importantly, spend some quality time working with VisiData.

Sit at the table

Some distributions, such as Arch Linux, carry VisiData in their software repositories. However, the quickest way to install the latest version is to use Pip, Python’s package manager.

If you already have Pip installed on your distro, run the pip3 install visidata command. You can check the version of VisiData with the vd --version command. The project is constantly under development and this tutorial is based on version 2.10.2.

Ubuntu and Fedora users can install Pip using the software repositories by installing the python3-pip package with the sudo apt install python3-pip or sudo dnf install python3-pip respectively.