How to Debug Python Code in Google Colab (Step-by-Step Guide)

Sharing is caring!

Introduction: How to Debug Python Code in Google Colab

If you use Colab daily, sooner or later you’ll run into errors — syntax mistakes, runtime issues, broken loops, wrong variable values. Knowing how to debug Python code in Google Colab is essential for fixing problems quickly and improving your workflow.

In this guide, you’ll learn multiple debugging methods: breakpoints, %debug, pdb, logging, error tracing, and recommended best practices. Everything is explained in a friendly, beginner-friendly way.

How to Debug Python Code in Google Colab (Full Methods Explained)

1. Use Colab’s Built-In Error Traceback

When a cell fails, Colab displays the full Python traceback.

Example

numbers = [1, 2, 3]
print(numbers[5])  # IndexError

How to read it

• Last line shows the error type (IndexError, KeyError, etc.)
• Above lines show the exact file and line number
• The topmost line of your code is usually where the real problem is

2. Use the %debug Magic Command in Google Colab

This is one of the fastest ways to debug.

Step-by-step

1. Run the cell that produces an error.
2. Immediately type in a new cell:

%debug

3. Press Enter.

What you can do inside %debug

• p variable → inspect value
• u → move up call stack
• d → move down call stack
• q → exit debugger

3. Use the Built-In Python Debugger (pdb) in Colab

If you need breakpoints inside your code, use pdb.

Example

import pdb

def calculate(a, b):
    pdb.set_trace()  # Breakpoint
    return a / b

calculate(10, 0)

Useful commands

4. Add Debugging Breakpoints with breakpoint()

Python 3.7+ supports built-in breakpoints.

Example

def example():
    x = 5
    breakpoint()
    return x

example()

5. Debug Python Code in Google Colab with Logging

Logging is better than print statements.

Example

import logging

logging.basicConfig(level=logging.INFO)

def multiply(a, b):
    logging.info(f"a={a}, b={b}")
    return a * b

multiply(10, 3)

Why logging is better

• Easier to toggle
• Cleaner output
• Useful for debugging loops

6. Debug Python Loops in Colab (Common Scenario)

Example

for i in range(5):
    print("Start loop:", i)
    breakpoint()

7. Use Colab’s Variable Inspector (Alternative)

Although not built-in like JupyterLab, you can install a variable inspector.

Install

!pip install colab-xterm

8. Use Try/Except Blocks for Controlled Debugging

Example

try:
    result = 10 / 0
except ZeroDivisionError as e:
    print("Error:", e)

Step-by-Step: How to Debug Python Code in Google Colab (Beginner Edition)

1. Run your script normally
2. Check the error message
3. Re-run the code with %debug
4. Inspect variables with p varname
5. Set breakpoints using breakpoint() or pdb.set_trace()
6. Use print/logging
7. Fix the code and re-run
8. Repeat

Troubleshooting & Common Debugging Problems in Colab

1. The debugger freezes or doesn’t exit

Type:

q

2. Logging messages don’t appear

Use:

logging.basicConfig(level=logging.DEBUG)

3. Colab runtime keeps restarting

Possible causes:

• Infinite loops
• Too much RAM usage
• Heavy GPU operations

4. “NameError: variable is not defined”

Restart runtime → Run all cells in order.

Best Practices for Debugging Python Code in Google Colab

• Run notebook top-to-bottom
• Keep functions small
• Add logging
• Use breakpoints for complex bugs
• Restart runtime when needed
• Use GitHub for backups
• Comment code fixes

Examples: Debugging Real Errors in Colab

Case 1: TypeError

a = "5"
b = 3
print(a + b)

Fix

print(int(a) + b)

Case 2: Infinite Loop

i = 0
while i < 10:
    print(i)
    # i += 1

FAQ (People Also Ask + Forums)

1. How do I debug line by line in Google Colab?

Use pdb.set_trace() or breakpoint().

2. Can I use VS Code–style debugging?

Not fully — but %debug and pdb are similar.

3. How do I inspect variables?

Inside debug mode:

p variable

4. Why does Colab skip my breakpoint?

Code block may not be executed.

5. Why does %debug say “No traceback available”?

You must trigger an error first.

6. Can I debug loops?

Yes — add breakpoint() inside the loop.

7. How do I restart debugging?

Runtime → Restart runtime.

8. Can I debug TensorFlow/PyTorch?

Yes — use logging or assertions.

9. How do I debug a function?

Insert breakpoint() inside it.

10. Why does my notebook crash?

Likely infinite loop or memory overload.

11. Is print debugging OK?

Yes — but logging is preferred.

12. What’s the easiest method?

%debug after an error.

Conclusion

Debugging is essential when working in Google Colab. With %debug, pdb, breakpoint(), logging, and tracebacks, you can quickly diagnose and fix issues.

👉 If this helped, bookmark the guide and share it!