Pandas替换列值并基于公式绘图，CSV按z分组绘制x/y与z的关系

Got it, let's break down these two Pandas data processing and visualization tasks with clear, actionable steps and code examples. I'll cover each requirement in detail below.

1. Replace Old Column Values & Visualize Derived New Columns #

First up: updating existing column values in a DataFrame and creating visualizations for the new columns you compute using your specified equations.

Step-by-Step Walkthrough #

• Import core libraries: You'll need pandas for data manipulation and matplotlib.pyplot (or seaborn) for plotting.
• Load your dataset: Use pd.read_csv() (or the right method for your data source) to pull in your data.
• Replace old values: Use replace() for simple value mappings, or apply() with a custom function if you need more complex logic.
• Calculate your new column: Plug your equation into a new column using Pandas vectorized operations (way faster than looping!).
• Visualize the results: Pick a plot type that highlights the relationship your new column shows—bar, line, scatter, etc.

Example Code #

import pandas as pd
import matplotlib.pyplot as plt

# Load your actual data (swap this path with your file)
df = pd.read_csv('your_dataset.csv')

# Example: Replace values in a 'status' column
value_map = {'inactive': 'archived', 'active': 'current'}
df['status'] = df['status'].replace(value_map)

# Example: Compute new column using an equation (adjust to your actual formula)
df['calculated_col'] = (df['col_a'] * 1.5) + (df['col_b'] ** 2)

# Plot: Average calculated value by updated status
plt.figure(figsize=(10, 6))
df.groupby('status')['calculated_col'].mean().plot(kind='bar', color='teal')
plt.title('Average Calculated Value by Status')
plt.xlabel('Status')
plt.ylabel('Average Calculated Value')
plt.xticks(rotation=0)
plt.tight_layout()
plt.show()

2. Group by Column z & Plot x/y Relationships #

For your CSV with z, x, y columns (where decimals use commas), we'll first fix the decimal parsing, then group by z to visualize how x and y interact across different z values.

Step-by-Step Walkthrough #

• Load data correctly: Use decimal=',' in pd.read_csv() to convert comma-separated decimals (like 1,75181E-07) into valid floats.
• Group by z: Split the data into subsets based on each unique z value.
• Plot for each group: Choose between subplots (to examine each z individually) or an overlay plot (to compare trends across z values).

Example Code #

import pandas as pd
import matplotlib.pyplot as plt

# Load CSV, handling comma decimals
df = pd.read_csv('your_zxy_data.csv', decimal=',')

# Option 1: Subplots for each unique z value
unique_z_values = df['z'].unique()
num_plots = len(unique_z_values)
fig, axes = plt.subplots(nrows=(num_plots + 1) // 2, ncols=2, figsize=(14, 8))
axes = axes.flatten()

for idx, z_val in enumerate(unique_z_values):
    group_data = df[df['z'] == z_val]
    axes[idx].scatter(group_data['x'], group_data['y'], color='coral', alpha=0.6)
    axes[idx].plot(group_data['x'], group_data['y'], color='navy', linewidth=1.5)
    axes[idx].set_title(f'Group: z = {z_val}')
    axes[idx].set_xlabel('x')
    axes[idx].set_ylabel('y')

# Hide empty subplots if needed
for ax in axes[num_plots:]:
    ax.axis('off')

plt.tight_layout()
plt.show()

# Option 2: Overlay all z groups in one plot for comparison
plt.figure(figsize=(10, 6))
for z_val, group in df.groupby('z'):
    plt.plot(group['x'], group['y'], label=f'z = {z_val}', linewidth=2)
plt.xlabel('x')
plt.ylabel('y')
plt.title('x vs y Across Different z Groups')
plt.legend()
plt.tight_layout()
plt.show()

Pro Tip #

Don't skip the decimal=',' parameter! Without it, Pandas will read your x and y values as strings instead of numbers, which breaks all your plotting and calculations.

内容的提问来源于stack exchange，提问作者Kumar