In my project, there are Chinese characters, and when I use the creat_plot() method to create a chart, the characters appear as garbage.

[superseanyoung]

datamapplot.create_plot(reduced_embeddings, doc_name)

[superseanyoung]

import matplotlib.pyplot as plt

设置中文字体

plt.rcParams['font.sans-serif'] = ['SimHei']

设置中文显示

plt.rcParams['axes.unicode_minus'] = False
Even using this code, I cannot solve the problem of garbled text.

[lmcinnes]

I'm afraid I can't be much help with this. The static plots rely on matplotlib's annotate function to render the text, so we are somewhat at the mercy of matplotlib's rendering. Ideally there would be a way to remedy this within matplotlib, but it looks like you've already explored the options there. If you can provide a minimal example (of just rendering some text in matplotlib using ax.annotate) with a garbled case, and some matplotlib that makes it "correct" I can possibly try to get fixes integrated into datamapplot, but if you don't know how to make matplotlib do the right thing then I don't think there is much I can do.

[NullPxl]

Hey @superseanyoung , I was having the same issue when my data contained many types of scripts. Here's how I solved it:

font_filepaths = [
    "/home/user/scratch/fonts/Noto_Sans/NotoSans-VariableFont_wdth,wght.ttf",
    "/home/user/scratch/fonts/Noto_Sans_Arabic/NotoSansArabic-VariableFont_wdth,wght.ttf",
    "/home/user/scratch/fonts/Noto_Sans_JP/NotoSansJP-VariableFont_wght.ttf",
    "/home/user/scratch/fonts/Noto_Sans_KR/NotoSansKR-VariableFont_wght.ttf",
    "/home/user/scratch/fonts/Noto_Sans_SC/NotoSansSC-VariableFont_wght.ttf",
    "/home/user/scratch/fonts/Noto_Sans_TC/NotoSansTC-VariableFont_wght.ttf"
]

font_families = [] # order of fallback
for font_path in font_filepaths:
    font = fm.FontProperties(fname=font_path)
    font_families.append(font.get_name())
    fm.fontManager.addfont(font_path)
    
fig,ax = datamapplot.create_plot(reduced_embeddings, custom_names,  font_family=font_families)

Hope this helps!

[DalelanW]

I believe to have found the easier fix to this. It seems specifying the font name explicitly worked

fig,ax = datamapplot.create_plot( reduced_embeddings, custom_names, font_family="Noto Sans JP")