importing adiftools changes/prevents notebook plotting in a kaggle notebook

[dan3dewey]

Hi,
I'm excited to see this project since I'm new to adif files, FT8, and PKSReporter.
I'm trying to use the PSKReporter adif files to see if it can help me understand/model my antenna pattern.

I'm using python in a kaggle notebook - the pip install works fine and so does "import adiftools.adiftools as adiftools" which lets me read the adif file into a dataframe - great!

But, the issue is that the "import ..." causes plotting in the notebook to stop working :(
Are there plot settings that the import changes that I can change back or reset?

Thanks,
-Dan

P.S. Here's code to show the problem in a kaggle python notebook:

import numpy as np
import matplotlib.pyplot as plt

import pandas as pd

!pip install -U adiftools

# Check plotting: Plot appears :)
plt.figure(figsize=(6,6))
plt.plot([1,2,3],[12,10,8])
plt.title("test")
plt.show()

import adiftools.adiftools as adiftools

# Check plotting again: No plot appears :(
plt.figure(figsize=(6,6))
plt.plot([1,2,3],[12,16,8])
plt.title("test")
plt.show()

[JS2IIU-MH]

Hello. Thank you for contacting me about the issue.
I was also able to reproduce the problem on Kaggle. As you pointed out, the plot was not displayed. Please give me some time to investigate where the problem is.
I apologize for noticing your message late. I appreciate you finding and using adiftools.

Thanks, JS2IIU

[JS2IIU-MH]

When adiftools is imported in a notebook environment such as Kaggle or Jupyter, matplotlib plots are no longer displayed with plt.show().

This appears to be caused by a global backend change during import.

Steps to reproduce:

import adiftools.adiftools as adiftools
import matplotlib.pyplot as plt

plt.figure(figsize=(6, 6))
plt.plot([1, 2, 3], [12, 16, 8])
plt.title("test")
plt.show()

Expected behavior:
Plots should be displayed normally in notebook environments.

Actual behavior:
No plot is displayed after importing adiftools.

Cause:
adiftools.adiftools imports adiftools.adifgraph, and adifgraph executes:

import matplotlib.pyplot as plt
matplotlib.use('Agg')

Because Agg is a non-interactive backend intended for file output, this overrides the notebook backend and prevents inline display in Kaggle/Jupyter.

Relevant code paths:

• adiftools.py imports adifgraph
• adifgraph.py forces matplotlib.use('Agg') at module import time

Why this is a problem:

• A library should generally not change the global matplotlib backend as an import side effect.
• This affects unrelated user plotting code, not only adiftools plotting functions.
• In notebook environments, users expect the active inline backend to remain intact.

Suggested fix:

• Remove the global matplotlib.use('Agg') from module import time.
• If Agg is needed for file export, enable it only in a dedicated script/CLI context, or make it optional.
• Prefer not to modify the global backend inside a reusable library.

Additional note:
The plotting functions already save figures with savefig(), so forcing Agg globally may not be necessary.

[JS2IIU-MH]

@dan3dewey

Hello,
The issue has now been resolved. I’ve released a new version, v0.1.11, so please give it a try. I also confirmed that the plot displays correctly on Kaggle. Thanks for your patience!