如何用Matplotlib绘制含年份、GDP、自杀数的Pandas数据集？

Let's break down what's going wrong here and fix your visualizations step by step:

First, the core issue: your suicides_gdp is a MultiIndex Series—its index has two levels (year and gdp_for_year ($)), with suicides_no as the values. Your original code fails because matplotlib can't handle MultiIndex data directly for axes, and you mixed up parameters for barh.

Step 1: Clean and Restructure the Data #

First, convert the MultiIndex Series to a flat DataFrame, and clean the GDP values (they're strings with commas, so we need to turn them into numeric values):

import pandas as pd
import matplotlib.pyplot as plt

# Convert the Series to a DataFrame and reset the index
suicides_gdp_df = suicides_gdp.reset_index()

# Clean GDP column: remove commas and convert to float
suicides_gdp_df[' gdp_for_year ($) '] = suicides_gdp_df[' gdp_for_year ($) '].str.replace(',', '').astype(float)

Step 2: Line Plot (Dual Y-Axis for Different Scales) #

Since GDP and suicide counts are on wildly different scales, we need a dual y-axis to visualize both clearly:

fig, ax1 = plt.subplots(figsize=(12, 6))

# Left Y-axis: Suicide Counts
color = 'tab:red'
ax1.set_xlabel('Year')
ax1.set_ylabel('Number of Suicides', color=color)
ax1.plot(suicides_gdp_df['year'], suicides_gdp_df['suicides_no'], color=color, marker='o', linewidth=2, label='Suicides')
ax1.tick_params(axis='y', labelcolor=color)

# Right Y-axis: GDP
ax2 = ax1.twinx()
color = 'tab:blue'
ax2.set_ylabel('GDP (USD)', color=color)
ax2.plot(suicides_gdp_df['year'], suicides_gdp_df[' gdp_for_year ($) '], color=color, marker='s', linewidth=2, label='GDP')
ax2.tick_params(axis='y', labelcolor=color)

# Add legend and title
fig.tight_layout()
fig.legend(loc='upper right', bbox_to_anchor=(1.15, 1))
plt.title('Russia: Suicide Counts vs GDP (1989-2015)')
plt.show()

Why your original line plot failed: You tried passing the entire MultiIndex (suicides_gdp.index) as the x-axis. Matplotlib can't interpret a 2-level index as a sequential x-value—we need to extract just the year column as our x-axis.

Step 3: Bar Plot (Vertical or Horizontal) #

Vertical Bar Plot (Suicides as Bars, GDP as Line) #

This is a common way to show categorical data (years) with two metrics:

fig, ax1 = plt.subplots(figsize=(14, 7))

# Vertical bars for suicide counts
color = 'tab:red'
bars = ax1.bar(suicides_gdp_df['year'], suicides_gdp_df['suicides_no'], color=color, alpha=0.7, label='Suicides')
ax1.set_xlabel('Year')
ax1.set_ylabel('Number of Suicides', color=color)
ax1.tick_params(axis='y', labelcolor=color)
# Rotate year labels to avoid overlap
plt.xticks(rotation=45)

# Add value labels to bars
for bar in bars:
    height = bar.get_height()
    ax1.text(bar.get_x() + bar.get_width()/2., height + 1000,
             f'{int(height)}',
             ha='center', va='bottom')

# Right Y-axis for GDP line
ax2 = ax1.twinx()
color = 'tab:blue'
ax2.plot(suicides_gdp_df['year'], suicides_gdp_df[' gdp_for_year ($) '], color=color, marker='o', linewidth=2, label='GDP')
ax2.set_ylabel('GDP (USD)', color=color)
ax2.tick_params(axis='y', labelcolor=color)

# Add legend and title
fig.tight_layout()
fig.legend(loc='upper right', bbox_to_anchor=(1.18, 1))
plt.title('Russia: Suicide Counts and GDP by Year')
plt.show()

Horizontal Bar Plot #

Your original barh code had parameter mix-ups—barh uses y for the categorical axis (years) and width for the numerical value (suicides). Here's the corrected version:

fig, ax1 = plt.subplots(figsize=(10, 12))

# Horizontal bars for suicide counts
color = 'tab:red'
bars = ax1.barh(suicides_gdp_df['year'], suicides_gdp_df['suicides_no'], color=color, alpha=0.7, label='Suicides')
ax1.set_ylabel('Year')
ax1.set_xlabel('Number of Suicides', color=color)
ax1.tick_params(axis='x', labelcolor=color)

# Add value labels to bars
for bar in bars:
    width = bar.get_width()
    ax1.text(width + 500, bar.get_y() + bar.get_height()/2,
             f'{int(width)}',
             ha='left', va='center')

# Top X-axis for GDP line (since it's a horizontal plot, use twiny())
ax2 = ax1.twiny()
color = 'tab:blue'
ax2.plot(suicides_gdp_df[' gdp_for_year ($) '], suicides_gdp_df['year'], color=color, marker='s', linewidth=2, label='GDP')
ax2.set_xlabel('GDP (USD)', color=color)
ax2.tick_params(axis='x', labelcolor=color)

# Add legend and title
fig.tight_layout()
fig.legend(loc='upper right', bbox_to_anchor=(1.2, 1))
plt.title('Russia: Suicide Counts and GDP by Year (Horizontal)')
plt.show()

Why your original barh failed: You passed x="suicides_no" and y=["year", " gdp_for_year ($) "]—this is backwards (barh uses y for categories) and you tried to pass two columns as the y-axis, which matplotlib can't handle. We only need year as the y-axis for horizontal bars.

内容的提问来源于stack exchange，提问作者A A