在 Streamlit 中部署数据帧的颜色条件格式

Question

执行代码后，转置列中的多个条目将有条件地突出显示。

我只想突出显示每个转置列中的顶部值和底部值。

以下是参考代码：

import pandas as pd
import streamlit as st

# Sample DataFrame
data = {
    'E0': ["Header1", "Header2", "Header3", "Header4"],
    'E1': [10, 20, 3, 5],
    'E2': [25, None, 20, 1],
    'E3': [5, 10, 8, None]
}
df = pd.DataFrame(data)

df_transposed = df.transpose()
custom_text = 'Not Available'
df_filled = df_transposed.fillna(custom_text)

def highlight_top(val):
    if val == df_transposed.loc['E1'].max() or val == df_transposed.loc['E2'].max() or val == df_transposed.loc['E3'].max():
        return 'background-color: green; color: white;'
    else:
        return ''

def highlight_bot(val):
    if val == df_transposed.loc['E1'].min() or val == df_transposed.loc['E2'].min() or val == df_transposed.loc['E3'].min():
        return 'background-color: red; color: white;'
    else:
        return ''

# Applying the style to the transposed DataFrame
styled_df = df_filled.style.applymap(highlight_top).applymap(highlight_bot)
st.table(styled_df)

输出表：

Answer 1

我会使用

describe

来获取

min

/

max

，然后使用

apply

预定义的 CSS 样式：

CSS = "background-color:{bc}; color:{c}" # add more if needed

STATS = {"min": {"stat": "min", "bg": "red", "c": "white"},
         "max": {"stat": "max", "bg": "green", "c": "white"}}

def highlight_mm(df_tos, stat, bg, c):
    # global df
    import numpy as np
    desc = df.drop(columns="E0").describe()
    mask = df_tos.eq(desc.loc[stat], axis=0)
    return np.where(mask, CSS.format(bc=bg, c=c), "")

styled_df = (
    df_filled.style
        .apply(highlight_mm, **STATS["min"], axis=None)
        .apply(highlight_mm, **STATS["max"], axis=None)    
        .format(precision=2) # this one is optional
)

st.table(styled_df)

输出（在 Streamlit）：