使用Matplotlib绘制Revenue柱状图不符合预期，寻求解决方案

Let's break down what's going wrong with your current code and how to adjust it to match your expected chart style:

Issues with Your Current Code #

• Incorrect bar width: Setting width=30 when your index is datetime-based means each bar spans 30 days, which causes overlapping and the squished, unreadable look you're seeing.
• Subplot layout: Your subplot2grid setup with rowspan=2 in a (6,7) grid combined with a tall figsize=(10,15) stretches the bars vertically in an unnatural way.
• Unformatted x-axis: Without proper date formatting, the x-axis labels become messy and unparseable.

Corrected Code #

First, ensure your DataFrame index is properly parsed as datetime (skip this step if it's already a datetime type):

import pandas as pd
import matplotlib.pyplot as plt
import matplotlib.dates as mdates

# Convert index to datetime if needed
df.index = pd.to_datetime(df.index)

# Adjust figure size and subplot layout for better proportions
fig = plt.figure(figsize=(12, 6))
ax1 = plt.subplot2grid((1,1), (0,0))  # Simplify to a single subplot for clarity

# Plot bars with appropriate width (matches daily interval here)
ax1.bar(df.index, df['revenue'], width=0.8, label='Revenue', align='center')

# Format x-axis dates to be readable
ax1.xaxis.set_major_locator(mdates.MonthLocator())  # Show monthly tick markers
ax1.xaxis.set_major_formatter(mdates.DateFormatter('%Y-%m'))  # Format labels as Year-Month
plt.xticks(rotation=45)  # Rotate labels to prevent overlap

# Add labels and legend for clarity
ax1.set_xlabel('Date')
ax1.set_ylabel('Revenue')
ax1.set_title('Revenue Over Time')
ax1.legend()

plt.tight_layout()  # Adjust layout to fit all labels and elements
plt.show()

Key Adjustments Explained #

• Bar width: width=0.8 works well for daily data (adjust to 25-28 if you want monthly bars, matching the average number of days in a month).
• Subplot layout: Using a simple 1x1 grid removes unnecessary vertical stretching, making the bars look natural and proportional.
• Date formatting: MonthLocator and DateFormatter clean up the x-axis, and rotating ticks ensures labels don't overlap and are easy to read.
• Layout adjustment: tight_layout() ensures all labels, titles, and chart elements fit properly within the figure boundaries.

If your data is aggregated by month instead of day, you can simplify further by plotting against month names/values directly instead of datetime indexes, but the above code works seamlessly for both daily and monthly datetime-indexed data.

内容的提问来源于stack exchange，提问作者Duong Bui