Optimization callbacks and monitoring in SciPy

Introduction

Callbacks help you watch and control the progress of an optimization. They let you see how the solution improves step-by-step.

You want to see how close the optimizer is to the best answer during the run.

You need to stop the optimization early if it is taking too long or not improving.

You want to log or print the current solution at each step for debugging.

You want to update a graph or progress bar while the optimizer runs.

Syntax

SciPy

def callback(xk):
    # xk is the current solution
    # Add your monitoring or stopping logic here
    pass

result = scipy.optimize.minimize(fun, x0, callback=callback)

The callback function is called after each iteration with the current solution.

You can use callback to print, log, or stop the optimization.

Examples

This prints the current solution at each step.

SciPy

def callback(xk):
    print(f"Current solution: {xk}")

This example shows how you might stop the optimization early based on a condition.

SciPy

def callback(xk):
    if some_condition(xk):
        print("Stopping early")
        return True  # Some optimizers support this to stop early

This saves each step's solution to a list for later analysis.

SciPy

def callback(xk):
    history.append(xk.copy())

Sample Program

This program minimizes a simple function. It prints and saves each step's solution using a callback.

SciPy

import numpy as np
from scipy.optimize import minimize

# Objective function: simple quadratic
fun = lambda x: (x[0]-3)**2 + (x[1]+1)**2

history = []

def callback(xk):
    print(f"Step solution: {xk}")
    history.append(xk.copy())

x0 = np.array([0, 0])
result = minimize(fun, x0, callback=callback)

print(f"Final solution: {result.x}")

Important Notes

Not all optimizers support stopping early via callback return values.

Callbacks run often, so keep them fast to avoid slowing optimization.

Use callbacks to collect data for graphs or logs to understand optimization behavior.

Summary

Callbacks let you watch optimization progress step-by-step.

You can print, log, or stop optimization using callbacks.

Callbacks help you understand and control how optimization runs.

Practice

1. What is the main purpose of using a callback function during optimization in scipy.optimize?

easy

A. To monitor and control the optimization process step-by-step

B. To automatically fix errors in the optimization function

C. To speed up the optimization by parallel processing

D. To save the final result to a file

Solution

1. Step 1: Understand the role of callbacks in optimization

   Callbacks are functions called at each iteration to observe or influence the optimization process.

2. Step 2: Identify the main use of callbacks

   They allow monitoring progress, logging, or stopping optimization early based on conditions.

3. Final Answer:

   To monitor and control the optimization process step-by-step -> Option A

4. Quick Check:

   Callbacks = monitor/control optimization [OK]

Hint: Callbacks watch optimization progress stepwise [OK]

Common Mistakes:

• Thinking callbacks fix errors automatically
• Assuming callbacks speed up optimization
• Believing callbacks save results directly

2. Which of the following is the correct way to define a callback function for scipy.optimize.minimize that prints the current parameter values at each iteration?

easy

A. def callback(): print(f"Current params: {xk}")

B. def callback(xk): return xk

C. def callback(xk, fk): print(f"Current params: {fk}")

D. def callback(xk): print(f"Current params: {xk}")

Solution

1. Step 1: Check callback function signature for scipy.optimize.minimize

   The callback receives one argument: the current parameter vector xk.

2. Step 2: Verify the function prints current parameters correctly

   def callback(xk): print(f"Current params: {xk}") defines callback(xk) and prints xk, which is correct.

3. Final Answer:

   def callback(xk): print(f"Current params: {xk}") -> Option D

4. Quick Check:

   Callback signature = one argument (xk) [OK]

Hint: Callback gets current params as single argument [OK]

Common Mistakes:

• Defining callback without parameters
• Using wrong parameter names or extra parameters
• Returning values instead of printing

3. Given the following code snippet, what will be printed during the optimization?

from scipy.optimize import minimize

def callback(xk):
    print(f"Step: {xk[0]:.2f}, {xk[1]:.2f}")

def func(x):
    return (x[0]-1)**2 + (x[1]-2)**2

res = minimize(func, [0, 0], callback=callback)

medium

A. Multiple lines showing parameter values approaching (1.00, 2.00)

B. Only one line showing initial parameters [0.00, 0.00]

C. No output because callback is ignored

D. Error because callback function has wrong signature

Solution

1. Step 1: Understand the callback usage in minimize

   The callback prints the current parameters at each iteration during optimization.

2. Step 2: Analyze the expected output

   Since the function minimizes distance to (1,2), parameters will update stepwise, printing multiple lines approaching (1.00, 2.00).

3. Final Answer:

   Multiple lines showing parameter values approaching (1.00, 2.00) -> Option A

4. Quick Check:

   Callback prints each step params [OK]

Hint: Callback prints each iteration's parameters [OK]

Common Mistakes:

• Assuming callback prints only once
• Thinking callback is ignored by minimize
• Believing callback signature is incorrect here

4. You wrote this callback to stop optimization early when the first parameter exceeds 0.5:

def callback(xk):
    if xk[0] > 0.5:
        return True

But the optimization does not stop early. What is the likely problem?

medium

A. The callback must raise an exception to stop optimization

B. The callback function must return False to stop optimization

C. Returning True does not stop optimization in scipy.optimize.minimize

D. The callback must be passed as a keyword argument named 'stop_callback'

Solution

1. Step 1: Understand callback behavior in scipy.optimize.minimize

   Callbacks can monitor progress but returning True does not stop the optimization.

2. Step 2: Identify correct way to stop optimization early

   To stop early, you must raise an exception or use other control mechanisms; returning True is ignored.

3. Final Answer:

   Returning True does not stop optimization in scipy.optimize.minimize -> Option C

4. Quick Check:

   Return True ≠ stop optimization [OK]

Hint: Returning True in callback won't stop optimization [OK]

Common Mistakes:

• Expecting return True to stop optimization
• Using wrong callback argument name
• Not raising exception to stop optimization

5. You want to log the optimization progress to a list and stop optimization if the function value goes below 0.01. Which callback implementation correctly achieves this with scipy.optimize.minimize?

hard

A. log = []
def callback(xk):
val = (xk[0]-1)**2 + (xk[1]-2)**2
log.append(val)
if val < 0.01:
return True

res = minimize(lambda x: (x[0]-1)**2 + (x[1]-2)**2, [0,0], callback=callback)

B. log = []
def callback(xk):
val = (xk[0]-1)**2 + (xk[1]-2)**2
log.append(val)
if val < 0.01:
raise StopIteration

res = minimize(lambda x: (x[0]-1)**2 + (x[1]-2)**2, [0,0], callback=callback)

C. log = []
def callback(xk, fk):
log.append(fk)
if fk < 0.01:
raise StopIteration

res = minimize(lambda x: (x[0]-1)**2 + (x[1]-2)**2, [0,0], callback=callback)

D. log = []
def callback(xk):
val = (xk[0]-1)**2 + (xk[1]-2)**2
log.append(val)
if val < 0.01:
return False

res = minimize(lambda x: (x[0]-1)**2 + (x[1]-2)**2, [0,0], callback=callback)

Solution

1. Step 1: Understand callback signature and logging

   The callback receives current parameters xk. We compute function value manually and append to log.

2. Step 2: Stopping optimization early

   Returning True or False does not stop optimization; raising an exception like StopIteration is a valid way.

3. Step 3: Verify options

   log = [] def callback(xk): val = (xk[0]-1)**2 + (xk[1]-2)**2 log.append(val) if val < 0.01: raise StopIteration res = minimize(lambda x: (x[0]-1)**2 + (x[1]-2)**2, [0,0], callback=callback) correctly logs values and raises StopIteration when value < 0.01. log = [] def callback(xk): val = (xk[0]-1)**2 + (xk[1]-2)**2 log.append(val) if val < 0.01: return True res = minimize(lambda x: (x[0]-1)**2 + (x[1]-2)**2, [0,0], callback=callback) returns True (ignored). log = [] def callback(xk, fk): log.append(fk) if fk < 0.01: raise StopIteration res = minimize(lambda x: (x[0]-1)**2 + (x[1]-2)**2, [0,0], callback=callback) has wrong callback signature (two args). log = [] def callback(xk): val = (xk[0]-1)**2 + (xk[1]-2)**2 log.append(val) if val < 0.01: return False res = minimize(lambda x: (x[0]-1)**2 + (x[1]-2)**2, [0,0], callback=callback) returns False (ignored).

4. Final Answer:

   log = [] def callback(xk): val = (xk[0]-1)**2 + (xk[1]-2)**2 log.append(val) if val < 0.01: raise StopIteration res = minimize(lambda x: (x[0]-1)**2 + (x[1]-2)**2, [0,0], callback=callback) -> Option B

5. Quick Check:

   Raise exception to stop + log values [OK]

Hint: Raise exception in callback to stop optimization [OK]

Common Mistakes:

• Returning True or False expecting to stop optimization
• Using wrong callback signature with two arguments
• Not computing function value inside callback