如何使用给定三列数据制作颜色编码散点图？

Got it, let's tackle this visualization request. First, I notice the original data you shared has mismatched lengths—member_birth_year has 13 entries, but member_gender and user_type only have 3 and 2 respectively. I'll start by generating a properly aligned synthetic dataset that uses your specified categories, then walk through two straightforward ways to build the color-coded scatter plot.

Step 1: Prepare Aligned Data #

First, let's create a consistent dataset where every birth year has a corresponding gender and user type:

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns

# Use your birth year list, then generate matching gender/user type values
birth_years = [1890, 1900, 1910, 1920, 1930, 1940, 1950, 1960, 1970, 1980, 1990, 2000, 2010]
genders = np.random.choice(['male','female', 'other'], size=len(birth_years))
user_types = np.random.choice(['customer', 'subscriber'], size=len(birth_years))

# Package into a DataFrame for easy handling
data = pd.DataFrame({
    'member_birth_year': birth_years,
    'member_gender': genders,
    'user_type': user_types
})

Option 1: Matplotlib (Full Control Over Colors) #

This approach lets you define exact colors for each category:

plt.figure(figsize=(10,6))

# Define custom color mapping for user types
color_map = {'customer': '#2ca02c', 'subscriber': '#ff9896'}

# Plot each user type group with its own color
for user_type, group in data.groupby('user_type'):
    # Add random jitter to y-axis to avoid overlapping points
    plt.scatter(group['member_birth_year'], np.random.normal(0, 0.4, size=len(group)),
                label=user_type, color=color_map[user_type], alpha=0.8, s=90)

# Customize plot labels and styling
plt.xlabel('Member Birth Year', fontsize=12)
plt.ylabel('Jittered Y-Axis (for readability)', fontsize=12)
plt.title('Birth Year vs User Type (Color-Coded)', fontsize=14)
plt.legend(title='User Type')
plt.grid(alpha=0.3)
plt.show()

Option 2: Seaborn (Simpler Categorical Coloring) #

Seaborn automatically handles categorical color mapping, which cuts down on boilerplate code:

plt.figure(figsize=(10,6))

# Scatter plot with color coded by member_gender (swap to 'user_type' if needed)
sns.scatterplot(data=data, x='member_birth_year', y=np.random.normal(0, 0.4, size=len(data)),
                hue='member_gender', palette='coolwarm', alpha=0.8, s=90)

# Polish the plot
plt.xlabel('Member Birth Year', fontsize=12)
plt.ylabel('Jittered Y-Axis (for readability)', fontsize=12)
plt.title('Birth Year vs Gender (Color-Coded)', fontsize=14)
plt.grid(alpha=0.3)
plt.show()

Quick Notes: #

• The random jitter on the y-axis is crucial here—without it, all points would stack on top of each other since birth years are discrete. If you have a real numerical y-variable, just replace np.random.normal(0, 0.4, ...) with that column.
• To switch which category defines the color, just swap hue='member_gender' (Seaborn) or the groupby key (Matplotlib) to user_type.
• Adjust s to change point size, alpha to tweak transparency, and palette/color_map to pick your preferred color scheme.

内容的提问来源于stack exchange，提问作者Rick J