I want to plot a matplotlib figure in a python virtual environment located in a tmpfs folder that was created by uv but can't.

This is what I did:

Step 1:

$ mkdir /dev/shm/test
$ cd /dev/shm/test
$ uv init --python 3.13 --cache-dir .cache
$ uv add matplotlib

Note: I used --cache-dir .cache to resolve the cache issue related to using uv to add package(s) in a folder that is located in a filesystem different to the filesystem where uv is installed in.

It is important for performance for the cache directory to be located on the same file system as the Python environment uv is operating on. Otherwise, uv will not be able to link files from the cache into the environment and will instead need to fallback to slow copy operations.

Step 2: I create a python script with path /dev/shm/test/main.py. It contains:

import matplotlib.pyplot as plt


def main():
    plt.figure()
    x = [2, 4, 6, 8]
    y = [10, 3, 20, 4]
    plt.plot(x, y)
    plt.show()


if __name__ == "__main__":
    main()

Step 3: Run main.py from within virtual environment in the tmpfs/ramdisk.

$ source .venv/bin/activate
(test) $ python main.py
/dev/shm/test/main.py:15: UserWarning: FigureCanvasAgg is non-interactive, and thus cannot be shown
  plt.show()

The matplotlib figure just won't plot. I don't know how to resolve this issue.

Update:

Per @furas comments, I tried these approaches:

$ uv add PyQt5  --cache-dir .cache

and

$ uv add PyQt6 --cache-dir .cache

separately. However, they still can't resolve the issue.

(test) $ python main.py
qt.qpa.plugin: Could not load the Qt platform plugin "xcb" in "" even though it was found.
This application failed to start because no Qt platform plugin could be initialized. Reinstalling the application may fix this problem.

Available platform plugins are: eglfs, linuxfb, minimal, minimalegl, offscreen, vnc, wayland-egl, wayland, wayland-xcomposite-egl, wayland-xcomposite-glx, webgl, xcb.

Aborted (core dumped)

I also noticed that tkinter won't work using the uv installed python.

(test) $ python -m tkinter
[xcb] Unknown sequence number while appending request
[xcb] You called XInitThreads, this is not your fault
[xcb] Aborting, sorry about that.
python: ../../src/xcb_io.c:157: append_pending_request: Assertion `!xcb_xlib_unknown_seq_number' failed.
Aborted (core dumped)

I want to plot a matplotlib figure in a python virtual environment located in a tmpfs folder that was created by uv but can't.

This is what I did:

Step 1:

$ mkdir /dev/shm/test
$ cd /dev/shm/test
$ uv init --python 3.13 --cache-dir .cache
$ uv add matplotlib

Note: I used --cache-dir .cache to resolve the cache issue related to using uv to add package(s) in a folder that is located in a filesystem different to the filesystem where uv is installed in.

It is important for performance for the cache directory to be located on the same file system as the Python environment uv is operating on. Otherwise, uv will not be able to link files from the cache into the environment and will instead need to fallback to slow copy operations.

Step 2: I create a python script with path /dev/shm/test/main.py. It contains:

import matplotlib.pyplot as plt


def main():
    plt.figure()
    x = [2, 4, 6, 8]
    y = [10, 3, 20, 4]
    plt.plot(x, y)
    plt.show()


if __name__ == "__main__":
    main()

Step 3: Run main.py from within virtual environment in the tmpfs/ramdisk.

$ source .venv/bin/activate
(test) $ python main.py
/dev/shm/test/main.py:15: UserWarning: FigureCanvasAgg is non-interactive, and thus cannot be shown
  plt.show()

The matplotlib figure just won't plot. I don't know how to resolve this issue.

Update:

Per @furas comments, I tried these approaches:

$ uv add PyQt5  --cache-dir .cache

and

$ uv add PyQt6 --cache-dir .cache

separately. However, they still can't resolve the issue.

(test) $ python main.py
qt.qpa.plugin: Could not load the Qt platform plugin "xcb" in "" even though it was found.
This application failed to start because no Qt platform plugin could be initialized. Reinstalling the application may fix this problem.

Available platform plugins are: eglfs, linuxfb, minimal, minimalegl, offscreen, vnc, wayland-egl, wayland, wayland-xcomposite-egl, wayland-xcomposite-glx, webgl, xcb.

Aborted (core dumped)

I also noticed that tkinter won't work using the uv installed python.

(test) $ python -m tkinter
[xcb] Unknown sequence number while appending request
[xcb] You called XInitThreads, this is not your fault
[xcb] Aborting, sorry about that.
python: ../../src/xcb_io.c:157: append_pending_request: Assertion `!xcb_xlib_unknown_seq_number' failed.
Aborted (core dumped)

• python
• matplotlib
• ramdisk
• tmpfs
• uv

• do you run GUI (X11/Wayland) on this system? As for me tmpfs shouldn't make problem but rather is problem with access graphic mode in system. – furas Commented Apr 1 at 11:00
• did you check UserWarning: FigureCanvasAgg is non-interactive, and thus cannot be shown plt.show() in Google? Maybe someone already had this problem. – furas Commented Apr 1 at 11:01
• python - UserWarning: FigureCanvasAgg is non-interactive, and thus cannot be shown plt.show() - Stack Overflow It suggests that GUI framework was not installed - tkinter or PyQt – furas Commented Apr 1 at 11:02
• another link found with Google - this time it is official forum for Matplotlib: FigureCanvasAgg interactivity problem - Community - Matplotlib - it also suggests that it needs GUI framework. – furas Commented Apr 1 at 11:04
• @furas I am using x11. I tried adding PyQt5 and PyQt6 but either of the still won't work. – Sun Bear Commented Apr 1 at 11:13

1 Answer 1

Presently, for Ubuntu 24.04 with Linux kernel 6.11.0-21-generic and uv-0.6.10 and uv-0.6.11, when I used uv to initiaIize a python virtual environment and then add the latest matplotlib version 3.10.1, the matplotlib package can show a plotted figure if Python 3.12 is initialized but not if Python 3.13 is initialized.

Works:

$ cd /dev/shm/
$ uv init test --python 3.12 --cache-dir test/.cache
$ cd test
$ uv add matplotlib --cache-dir .cache
$ cp ~/main.py ./   # copy the file mentioned in question to this folder
$ uv run main.py  

Don't Work:

$ cd /dev/shm/
$ uv init test --python 3.13 --cache-dir test/.cache
$ cd test
$ uv add matplotlib --cache-dir .cache
$ cp ~/main.py ./   # copy the file mentioned in question to this folder
$ uv run main.py
/dev/shm/test/main.py:9: UserWarning: FigureCanvasAgg is non-interactive, and thus cannot be shown
  plt.show()

I have also tried switching from Python 3.13 to Python 3.12 and then run the main.py and that works.

$ nano pyproject.toml
# Amend line `requires-python = ">=3.13"` to `requires-python = ">=3.12"`
$ uv venv --python 3.12 --cache-dir .cache
$ uv run main.py

Looking at uv's documentation, it seems that Python 3.13 is not recommended for use yet, although uv can download and install it to ~/.local/share/uv/python/cpython-3.13.2-linux-x86_64-gnu/bin/python3.13.