jubujjamer/npy_viewer

A bunch of utilities to quickly visualize .npy files

npy_viewer

A bunch of utilities to quickly visualize .npy files

Information

This repository contains pairs of .desktop files and Python scripts, that allow to open .npy files with a double click and quickly visualize the array that it contains, as well as statistics about it. The scripts also work when opening multiple files at once.

The .desktop files follow the Desktop Entry specifications, so they should work for most Linux. It was tested on Ubuntu 16.04.

Requirements

The scripts are written in Python 3. The .desktop files will use the python that python3 refers to in your system.

To change that, simply update the path in the .desktop files. For example: sed -i 's,python3,/usr/local/bin/python3,' NPY*.desktop

Packages

Additionally to numpy and matplotlib, some packages are required for certain scripts:

• NPYCopyAsCSV: pyperclip csv

Backends

Depending on you system, there might or might not be an exisiting backend that matplotlib can use. It if doesn't find one, it should tell you what to install.

For NPY3DArrayViewer, the TkAgg backend is a little shaky when changing files, so you might want to use a Qt backend, such as Qt4Agg or Qt5Agg.

The TkAgg backend requires the deb package python3-tk. The Qt4Agg backend requires the deb package python3-pyqt4

Installing

After downloading the repository or indivual files to your computer, the paths to the Python scripts need to be updated in the .desktop files.

You can do so:

• Automagically, by running setup_npy_scripts.sh, which will put them in ~/.local/share/applications
• Manually by changing the lines in the .desktop files. Exec= python3 /path/to/python/script.py %U

On Ubuntu, the usual directories to put .dekstop files in are :

• /usr/share/applications/ for all users to access (requires sudo privileges)
• ~/.local/share/applications/ for a specific user

After that, the application icon should appear in the Open With... menu. You can set it as default to open it with just a double click.

Have fun ! 😉