基于Plotly实现支持X轴滚动的分组柱状图技术求助

Hey there! Let's sort out your grouped bar chart issue—you're right, the code you tried is creating a stacked chart because you're feeding all your data into a single Bar trace. Here's how to fix it, add category-specific colors, and enable X-axis scrolling for all your dates:

Step 1: Preprocess Your Data First #

Your raw data has duplicate entries for the same category and date (like two apple entries in 10/2020). We need to aggregate these values first so each date-category pair has one total value. Using pandas' groupby makes this easy:

import pandas as pd

# Aggregate values by date and category, fill missing categories with 0
aggregated_df = df.groupby(['date', 'category'])['value'].sum().unstack(fill_value=0)

Step 2: Build the Grouped Bar Chart #

To get grouped bars, you need a separate Bar trace for each category. This lets Plotly display them side-by-side instead of stacking. We'll also assign custom colors to each category for clarity:

import plotly.graph_objects as go

# Define custom colors for each category (adjust these to your preference)
category_colors = {
    'apple': '#ff6b6b',
    'banana': '#4ecdc4',
    'strawberry': '#ffe66d'
}

# Initialize the figure
fig = go.Figure()

# Add a trace for each category
for category in aggregated_df.columns:
    fig.add_trace(go.Bar(
        x=aggregated_df.index,
        y=aggregated_df[category],
        name=category,
        text=category,
        textposition='auto',
        marker_color=category_colors.get(category, '#1a535c')  # Fallback color if category is new
    ))

Step 3: Enable X-Axis Scrolling & Final Layout Tweaks #

Add a rangeslider to the X-axis so you can scroll through all your dates, and set barmode='group' explicitly to ensure bars are grouped, not stacked:

fig.update_layout(
    barmode='group',
    title='Monthly Value by Category',
    xaxis=dict(
        title='Month/Year',
        rangeslider=dict(visible=True),  # Turn on the scrollable rangeslider
        type='category'  # Treat dates as categorical values (works for string dates like '10/2020')
    ),
    yaxis=dict(title='Total Value')
)

# Show the chart
fig.show()

Key Notes #

• Why separate traces? Plotly only groups bars when each category is its own data series. A single trace will always stack all values together.
• Handling new categories: The fallback color in marker_color ensures any new categories you add later will still get a consistent color without updating the code.
• Date format flexibility: If you switch your dates to a standard date format (e.g., '2020-10'), you can change xaxis.type to 'date' for more intelligent scrolling (like zooming by month/year).

内容的提问来源于stack exchange，提问作者mlenthusiast