[Python Tips] Streamlit: A Rapid Prototyping Tool for Python

algieba — Mon, 23 Dec 2024 16:32:58 +0000

This article aims to quickly get you started with the Python streamlit library, including installation, input data, chart drawing, basic controls, progress bars, and free deployment.

Streamlit, as the official website claims, helps you build and share data applications faster! It is well-suited for data processing, data presentation, machine learning prototypes, AI prototype demonstrations, and even simple web applications. With low difficulty in getting started, simple deployment, and free public access to Streamlit Sharing provided by the official platform.

Quick Introduction to the Features of Streamlit

High Integration with Data Science Tools. Libraries such as Matplotlib, Pandas, Plotly, Altair, etc., are well-adapted in streamlit, allowing direct display of charts and data generated by these libraries in applications, and Streamlit's presence can be seen in many related academic papers.
Easy Deployment. You can complete the deployment of applications with a simple command, running anywhere, locally, on cloud servers, or directly hosted on the official Streamlit Sharing.
Real-time Updates. Streamlit applications update in real-time, automatically refreshing whenever the data changes.

0x00 Installation

Ensure your Python version is above 3.8
Install directly with pip pip install streamlit OK, the installation is complete. You can use the command streamlit hello

Whoosh~ The built-in HelloWorld Demo of streamlit pops up, you can scroll through the demo on the left.

0x01 Before Coding

Before starting to code, let's talk about an important point -> how to start a streamlit app script

Unlike general Python scripts, which can be run with python xxx.py, streamlit requires python -m streamlit run xxx.py to run. (The script following streamlit run can even be a GitHub repository link, for example: streamlit run https://github.com/Algieba-dean/ZhuGeHorary/blob/master/ZhuGeHorary_APP.py)

Streamlit will monitor changes in real-time, and code changes will be synchronized to the deployed App upon saving.

0x02 Displaying Data

st.write()

st.write(), if you don't know what to use, just use it. Even with st.table(), st.dataframe() there are no special requirements.

import streamlit as st
import pandas as pd
st.write("Display table:")
st.write(pd.DataFrame({
    'Column A': [1, 2, 3, 4],
    'Column B': [101, 202, 303, 404]
}))

OK, now we've learned the simplest data presentation.

0x03 Input Control

st.text_input

The simplest input control, for text input.

text = st.text_input("Enter some text")
st.write(text)

OK, with these two, you can complete a large part of the tasks.

0x04 Advanced Controls

st.number_input for numerical input

number = st.number_input("Enter a number between 0-100", min_value=0, max_value=100, step=1)

st.slider for slider

slider_number = st.slider(f"Try sliding to the number: {number}", min_value=0, max_value=10000, value=[0, 500])

st.file_uploader for file uplaoding

uploaded_file = st.file_uploader("upload a data file", type=['csv', 'txt'])

0x05 Progress Bar

st.progress

import  time
st.write("Calculation simulation in progress...")

latest_iteration = st.empty()

bar = st.progress(0)

for i in range(100):
    latest_iteration.text(f"now it's {i+1}")
    bar.progress(i + 1)
    time.sleep(0.1)

st.write("Simulation calculation complete!")

0x06 Application Deployment

Upload your code to Github and set it as a public repository
Log in to streamlit share with your GitHub account
Click on Create app at the top right
Click on Deploy a public app from GitHub and then Deploy now
Fill in the repository information correctly, script filename, configure a usable URL
Click Deploy
After waiting, your app will be online

End

Welcome to leave your app link here

[Development Tips] How to Get the Number of Strokes in Chinese Characters?

algieba — Sun, 22 Dec 2024 14:55:06 +0000

Background

During the development of a simple divination script, I encountered this interesting problem. If it's just a specific few Chinese characters, we can hard-code a dictionary in the script, but what if we want to get the stroke count for any Chinese character?

pypinyin library

from pypinyin import pinyin, Style

def get_strokes_count(chinese_character):
    pinyin_list = pinyin(chinese_character, style=Style.NORMAL)
    strokes_count = len(pinyin_list[0])
    return strokes_count

character = input("Please enter a Chinese character：")
strokes = get_strokes_count(character)
print("Character'{}'stroke numbers：{}".format(character, strokes))

I tried it and found that the result is actually the number of results in the normal pinyin format for that character.

Unihan Database

The Unihan database is a Chinese character database maintained by the Unicode Consortium, which seems quite reliable and also provides online tools.

In its online query tooUnihan Database LookupI found that the query results contain the kTotalStrokes field, which is the stroke count data we need.
As the official database of Unicode, the current version fully meets the basic needs of Chinese character queries.

Nice! One step closer to success!

Getting Stroke Information from Unihan Database

I initially planned to send query requests directly through lookup, but it was too slow, and the address is abroad from China. I found that the database file itself is not large, so I downloaded it directly.

Unihan

AfterAfter opening the compressed package, there are several files.

By looking up the results, we need the kTotalStrokes field in the IRG Source. Extract this file.
I tested the regex on regex101 to extract the desired Unicode part and stroke count part, and saved them separately for querying.

Coding

Extracting Stroke Information

file = Path("Stroke/Unihan_IRGSources.txt")
output = Path("Stroke/unicode2stroke.json")
stroke_dict = dict()
with open(file,mode="r") as f:
    for line in f:
        raw_line = line.strip()
        pattern = r"(U\+.*)\skTotalStrokes.*\s(\d+)"
        result = re.findall(pattern=pattern, string=raw_line)
        if len(result) == 0:
            continue
        unicode_key = result[0][0]
        unicode_stroke = result[0][1]
        print(f"{unicode_key}: {unicode_stroke}")
        stroke_dict[unicode_key] = unicode_stroke

with open(file=output, mode="w", encoding="utf-8") as f:
    json.dump(stroke_dict,f, ensure_ascii=False, indent=4)

exported to json for easy access.

Writing the Acquisition Function

with open(output) as f:
    unicode2stroke = json.load(f)

def get_character_stroke_count(char: str):
    unicode = "U+" + str(hex(ord(char)))[2:].upper()
    return int(unicode2stroke[unicode])

test_char = "阿"
get_character_stroke_count(char=test_char)

When obtaining, note that Unicode converts the character to its corresponding hexadecimal code

Success! The expected result is achieved!

DEV Community: algieba