Poor Result when avoiding a polygonal curve about adjusttext

[Phlya]

Would you be able to share some example data?

[Phlya]

Also - are you using the latest version?

[ThomasParistech]

I'm using adjustText==1.1.1

Here's a minimal reproducible example (with JSON data)

import json
import matplotlib.pyplot as plt
from adjustText import adjust_text

path = "data/test.json"

with open(path, encoding="utf-8") as f:
    data = json.load(f)
    curve_x = data["curve_x"]
    curve_y = data["curve_y"]
    list_x = data["list_x"]
    list_y = data["list_y"]
    list_text = data["list_text"]

plt.plot(curve_x, curve_y, color="r", lw=2)
plt.plot(list_x, list_y, '^', color="g", markersize=10)


texts = [plt.text(x, y, txt, ha='center', va='center', fontsize=8)
         for x, y, txt in zip(list_x, list_y, list_text)]


adjust_text(texts,
            x=curve_x,
            y=curve_y,
            time_lim=2,
            arrowprops=dict(arrowstyle='->', color='red'))


plt.axis('equal')
plt.gca().invert_yaxis()

plt.show()

test.json

[ThomasParistech]

I got roughly the same result when removing plt.axis('equal') and plt.gca().invert_yaxis()

[Phlya]

When there are so many points it's a bit tricky. Here is some arguments in the right direction, you can experiment a bit more.

adjust_text(texts, x=curve_x, y=curve_y, force_explode=0, force_static=(2e-3, 5e-3),
            time_lim=2, expand=(1, 1),
            arrowprops=dict(arrowstyle="->", color='r', lw=0.5))

[ThomasParistech]

Thanks for the parameter suggestion!
But it turns out that textalloc works better in this specific scenario (ckjellson/textalloc#18), I'll use it instead
cheers :)