How can I print Japanese Characters via Python Visual?

[iamriz]

My data table contains one column of Japanese words, when I print this to a chart as data label using Python Visual, 3 rectangular characters ( ▯▯▯) will be printed instead. Kindly help how to print the JP characters properly. Thank you very much!

Excerpt of my Python Visual Script:

---

# dataset = pandas.DataFrame(1_ave, 1_slope, 顧客名)
# dataset = dataset.drop_duplicates()

# Paste or type your script code here:
import matplotlib.pyplot as plt
import json
import datetime
today = datetime.datetime.today()
current_month = today.month

fig, ax = plt.subplots(figsize=(18, 8)) # set chart size
ax.scatter(dataset["1_ave"], dataset["1_slope"], s=80, marker="s", color='r') #square
plt.xlabel('Average')
plt.ylabel('Slope')

ax.annotate("顧客名", (1, 1)) #Trying to print only one Japanese word here because putting all data will lead to script timeout

plt.show()

---

[v-yuta-msft · Accepted solution]

@iamriz ,

This is not the issue in power bi, if you create a dataframe and run the code this issue also occurs. You may haven't configured the matplotlibrc, please download the .ttf font file and move the file to matplotlib/mpl-data/fonts/ttf then configure font.family, font.sans-serif and axes.unicode_minus like below:

font.family : sans-serif        
font.sans-serif : add the ttf file name here   
axes.unicode_minus : False

Finally run code below to reload the python environment:

from matplotlib.font_manager import _rebuild

_rebuild() #reload一下

Community Support Team _ Jimmy Tao

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.

[iamriz]

Thanks, @v-yuta-msft, sorry, my mistake I didn't check with pure python app using matplotlib.

Adding 

plt.rcParams['font.sans-serif'] = ['MS Gothic', 'sans-serif']

worked for me.

[amitchandak]

@iamriz ,

Refer, if this can help https://docs.python.org/3/howto/unicode.html

[iamriz]

Hi @amitchandak , I already tried str.encode() which returns b'\xe9\xa1\xa7\xe5\xae\xa2\xe5\x90\x8d', then when I decode it again, it will print the rectangle(garbage) characters. Actually, I thought that I do not need encoding and decoding because in Power BI data model, the JP characters perfectly show in utf-8 format. But I don't understand why it doesnt show up properly with Python Visual. Any advice?