streamlit-example-vibe / src /streamlit_app.py
JUNGU's picture
Update src/streamlit_app.py
68fece7 verified
raw
history blame
4.37 kB
import streamlit as st
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
from scipy.stats import norm, skew
import platform
# ํ•œ๊ธ€ ํฐํŠธ ์„ค์ • (Windows, Mac, Linux ํ™˜๊ฒฝ์— ๋งž๊ฒŒ ์ž๋™ ์„ค์ •)
if platform.system() == 'Windows':
plt.rc('font', family='Malgun Gothic')
elif platform.system() == 'Darwin': # Mac
plt.rc('font', family='AppleGothic')
else: # Linux
# ๋‚˜๋ˆ”๊ณ ๋”• ํฐํŠธ๊ฐ€ ์„ค์น˜๋˜์–ด ์žˆ์–ด์•ผ ํ•ฉ๋‹ˆ๋‹ค.
# sudo apt-get install fonts-nanum*
plt.rc('font', family='NanumGothic')
plt.rcParams['axes.unicode_minus'] = False # ๋งˆ์ด๋„ˆ์Šค ํฐํŠธ ๊นจ์ง ๋ฐฉ์ง€
def main():
"""
์ŠคํŠธ๋ฆผ๋ฆฟ์„ ์ด์šฉํ•œ ํ•™์ƒ ์ ์ˆ˜ ๋ถ„ํฌ ๋ถ„์„ ์• ํ”Œ๋ฆฌ์ผ€์ด์…˜
"""
st.title("ํ•™์ƒ ์ ์ˆ˜ ๋ถ„ํฌ ๋ถ„์„ ๋„๊ตฌ ๐Ÿ“Š")
st.write("CSV ํŒŒ์ผ์„ ์—…๋กœ๋“œํ•˜์—ฌ ํ•™์ƒ๋“ค์˜ ์ ์ˆ˜ ๋ถ„ํฌ๋ฅผ ํ™•์ธํ•˜๊ณ , ์ •๊ทœ๋ถ„ํฌ์™€์˜ ์ฐจ์ด ๋ฐ ์™œ๋„(skewness)๋ฅผ ๋ถ„์„ํ•ฉ๋‹ˆ๋‹ค.")
st.write("---")
# ํŒŒ์ผ ์—…๋กœ๋“œ ์œ„์ ฏ
uploaded_file = st.file_uploader("์ ์ˆ˜ ๋ฐ์ดํ„ฐ๊ฐ€ ํฌํ•จ๋œ CSV ํŒŒ์ผ์„ ์—…๋กœ๋“œํ•˜์„ธ์š”.", type="csv")
if uploaded_file is not None:
try:
# utf-8-sig ์ธ์ฝ”๋”ฉ์œผ๋กœ CSV ํŒŒ์ผ ์ฝ๊ธฐ
df = pd.read_csv(uploaded_file, encoding='utf-8-sig')
st.subheader("์—…๋กœ๋“œ๋œ ๋ฐ์ดํ„ฐ ๋ฏธ๋ฆฌ๋ณด๊ธฐ")
st.dataframe(df.head())
# ๋ถ„์„ํ•  ์ ์ˆ˜ ์—ด ์„ ํƒ
score_column = st.selectbox("๋ถ„์„ํ•  ์ ์ˆ˜ ์—ด(column)์„ ์„ ํƒํ•˜์„ธ์š”:", df.columns)
if score_column:
# ์„ ํƒ๋œ ์—ด์˜ ๋ฐ์ดํ„ฐ ์ถ”์ถœ (๊ฒฐ์ธก์น˜ ์ œ๊ฑฐ)
scores = df[score_column].dropna()
if pd.api.types.is_numeric_dtype(scores):
st.subheader(f"'{score_column}' ์ ์ˆ˜ ๋ถ„ํฌ ๋ถ„์„ ๊ฒฐ๊ณผ")
# 1. ๊ธฐ์ˆ  ํ†ต๊ณ„๋Ÿ‰ ํ‘œ์‹œ
st.write("#### ๐Ÿ“ˆ ๊ธฐ์ˆ  ํ†ต๊ณ„๋Ÿ‰")
st.table(scores.describe())
# 2. ๋ถ„ํฌ ์‹œ๊ฐํ™” ๋ฐ ์ •๊ทœ๋ถ„ํฌ ๋น„๊ต
st.write("#### ๐ŸŽจ ์ ์ˆ˜ ๋ถ„ํฌ ์‹œ๊ฐํ™”")
fig, ax = plt.subplots(figsize=(10, 6))
# ํžˆ์Šคํ† ๊ทธ๋žจ ๋ฐ KDE ํ”Œ๋กฏ
sns.histplot(scores, kde=True, stat='density', label='ํ•™์ƒ ์ ์ˆ˜ ๋ถ„ํฌ', ax=ax)
# ์ •๊ทœ๋ถ„ํฌ ๊ณก์„  ์ถ”๊ฐ€
mu, std = norm.fit(scores)
xmin, xmax = plt.xlim()
x = np.linspace(xmin, xmax, 100)
p = norm.pdf(x, mu, std)
ax.plot(x, p, 'k', linewidth=2, label='์ •๊ทœ๋ถ„ํฌ ๊ณก์„ ')
title = f"'{score_column}' ์ ์ˆ˜ ๋ถ„ํฌ (ํ‰๊ท : {mu:.2f}, ํ‘œ์ค€ํŽธ์ฐจ: {std:.2f})"
ax.set_title(title)
ax.set_xlabel('์ ์ˆ˜')
ax.set_ylabel('๋ฐ€๋„')
ax.legend()
st.pyplot(fig)
# 3. ์™œ๋„(Skewness) ๊ณ„์‚ฐ ๋ฐ ํ•ด์„
st.write("#### ๐Ÿ“ ์™œ๋„ (Skewness) ๋ถ„์„")
skewness = skew(scores)
st.metric(label="์™œ๋„ (Skewness)", value=f"{skewness:.4f}")
if skewness > 0.5:
st.info("๊ผฌ๋ฆฌ๊ฐ€ ์˜ค๋ฅธ์ชฝ์œผ๋กœ ๊ธด ๋ถ„ํฌ (Positive Skew): ๋Œ€๋ถ€๋ถ„์˜ ํ•™์ƒ๋“ค์ด ํ‰๊ท ๋ณด๋‹ค ๋‚ฎ์€ ์ ์ˆ˜์— ๋ชฐ๋ ค์žˆ๊ณ , ์ผ๋ถ€ ํ•™์ƒ๋“ค์ด ๋งค์šฐ ๋†’์€ ์ ์ˆ˜๋ฅผ ๋ฐ›์•˜์Šต๋‹ˆ๋‹ค.")
elif skewness < -0.5:
st.info("๊ผฌ๋ฆฌ๊ฐ€ ์™ผ์ชฝ์œผ๋กœ ๊ธด ๋ถ„ํฌ (Negative Skew): ๋Œ€๋ถ€๋ถ„์˜ ํ•™์ƒ๋“ค์ด ํ‰๊ท ๋ณด๋‹ค ๋†’์€ ์ ์ˆ˜์— ๋ชฐ๋ ค์žˆ๊ณ , ์ผ๋ถ€ ํ•™์ƒ๋“ค์ด ๋งค์šฐ ๋‚ฎ์€ ์ ์ˆ˜๋ฅผ ๋ฐ›์•˜์Šต๋‹ˆ๋‹ค.")
else:
st.info("๋Œ€์นญ์— ๊ฐ€๊นŒ์šด ๋ถ„ํฌ: ์ ์ˆ˜๊ฐ€ ํ‰๊ท ์„ ์ค‘์‹ฌ์œผ๋กœ ๋น„๊ต์  ๊ณ ๋ฅด๊ฒŒ ๋ถ„ํฌ๋˜์–ด ์žˆ์Šต๋‹ˆ๋‹ค.")
else:
st.error(f"์˜ค๋ฅ˜: ์„ ํƒํ•˜์‹  '{score_column}' ์—ด์€ ์ˆซ์ž ๋ฐ์ดํ„ฐ๊ฐ€ ์•„๋‹™๋‹ˆ๋‹ค. ์ˆซ์ž ๋ฐ์ดํ„ฐ๋กœ ๊ตฌ์„ฑ๋œ ์—ด์„ ์„ ํƒํ•ด์ฃผ์„ธ์š”.")
except Exception as e:
st.error(f"ํŒŒ์ผ์„ ์ฝ๋Š” ๋„์ค‘ ์˜ค๋ฅ˜๊ฐ€ ๋ฐœ์ƒํ–ˆ์Šต๋‹ˆ๋‹ค: {e}")
st.warning("CSV ํŒŒ์ผ์ด 'utf-8-sig' ๋˜๋Š” 'utf-8' ์ธ์ฝ”๋”ฉ ํ˜•์‹์ธ์ง€ ํ™•์ธํ•ด์ฃผ์„ธ์š”.")
if __name__ == '__main__':
main()