My favorite place to debug a Python application is in a full IDE like VSCode or PyCharm. However, sometimes those options aren’t available, soit’s good to know the alternatives. This post shows some basic functionality of some Python debuggers.

Table of Contents

• pdb
  • Main commands
  • List all the attributes of an object
  • See the traceback
• ipdb
  • Context
• Jupyter notebooks
• Command line

pdb

pdb is the default Python debugger and for this reason alone it’s good to be familiar with. I often use it when I am tunneling into somewhere.

Main commands

The main commandes to know are:

• n - proceed to next step
• s - step into
• u - move up stack
• d - move down stack
• l - show surrounding lines
• ll - “long list”, show all of the code

List all the attributes of an object

p dir(a)

See the traceback

import traceback
traceback.print_stack()

ipdb

You can use the same commands as in pdb but there’s some extra functionality on top.

Context

Automatically see the context above and below the your current line. I think this feature alone makes it preferable to pdb

ipdb.set_trace(context=11)

Jupyter notebooks

If you have an error, you can recover the state with %debug.

Command line

You can also set the debugger to capture the traceback if you have an error in a command, right from the command line:

python -m pdb algo_trainer.py