Windows系统Scipy optimize导入错误：无法导入getargspec_no_self

Hey there! Let's tackle this error you're seeing with Scipy's optimize module on Windows. First, let's break down what's going on, then walk through solutions and fix the small bugs in your code too.

Why This Error Happens #

This import error almost always stems from version incompatibility between Scipy and its dependencies (like NumPy). The getargspec_no_self function was either removed or renamed in newer versions of Scipy, or your NumPy version is too old to work with your current Scipy release.

Step-by-Step Solutions for Windows #

Here are a few fixes you can try, starting with the most common:

1. Downgrade Scipy to a Compatible Version #

Newer Scipy versions (1.12.x+) removed this internal utility function. Try rolling back to a stable, compatible version like 1.11.4:
Open Command Prompt or PowerShell and run:

pip install scipy==1.11.4 --force-reinstall

The --force-reinstall flag ensures your existing Scipy installation is fully replaced.

2. Update NumPy to Match Your Scipy Version #

If downgrading Scipy isn't your preference, make sure your NumPy version is up-to-date and compatible:

pip install --upgrade numpy

3. Use a Fresh Virtual Environment (Best for Clean Slate) #

Sometimes existing environment conflicts cause weird issues. Create a new virtual environment to isolate your dependencies:

# Create a new virtual environment
python -m venv scipy_optimize_env
# Activate it (Windows only)
scipy_optimize_env\Scripts\activate
# Install the latest compatible packages
pip install scipy numpy matplotlib

Fixes for Your Code Snippets #

Even after resolving the import error, your code has small bugs that will cause issues—let's fix those too:

Scene 1 Correction #

You passed x to optimize.minimize instead of your function f:

import matplotlib.pyplot as plt
import numpy as np
from scipy import optimize

def f(x):
    return x**2 + 10 * np.sin(x)

x_data = np.arange(-10, 10, 0.1)
plt.plot(x_data, f(x_data))
plt.show()

# Fixed: pass the function `f` instead of `x`
result = optimize.minimize(f, x0=0)
print(result)

Scene 2 Correction #

You used sin without referencing NumPy (since you didn't import it separately):

import matplotlib.pyplot as plt
import numpy as np
from scipy import optimize

x_data = np.linspace(-5, 5, num=50)
y_data = 2.9 * np.sin(1.5 * x_data) + np.random.normal(size=50)

def test_func(x, a, b):
    # Fixed: use np.sin instead of sin
    return a * np.sin(b * x)

params, params_covariance = optimize.curve_fit(test_func, x_data, y_data, p0=[2,2])
print(params)

Give these steps a shot—they should resolve both the import error and the code bugs!

内容的提问来源于stack exchange，提问作者Jackipline