Quickstart

NB-QOL provides various utilities to enhance your Jupyter notebook workflow. Here are some common use cases:

CUDA Device Management

Check and manage your CUDA devices:

from nbqol import devices

# List all available CUDA devices
devices.cuda_device_report()

# Select specific CUDA devices to use
devices.set_cuda_visibles(0, 1)  # Use CUDA devices 0 and 1

IPython Configuration

Configure IPython environment:

import nbqol

# Enable autoreload to automatically reload modules during development
nbqol.set_autoreload('complete')

# Hide warning messages
nbqol.hide_warnings()

# Get environment settings
env_settings = nbqol.get_main_env_settings()

# Alternative: use the options module directly
from nbqol import options
options.set_autoreload('complete')

Path Operations

Work with notebook paths:

import nbqol

# Get the path of the current notebook
nb_path = nbqol.notebook_path()

# Find the Git repository root
git_root = nbqol.path_to_git_root()

# Add git root to Python path
nbqol.add_git_root_to_sys_path()

# Alternative: use the env module directly
from nbqol import env
nb_path = env.notebook_path()

Styling Notebooks

Apply CSS styling to your notebooks:

import nbqol

# Auto-detect environment and apply appropriate styling
nbqol.auto_style()

# Alternative: use the stylizer module directly
from nbqol import stylizer
stylizer.auto_style()

# Load specific VSCode styles
from nbqol.styles import load_vscode_styles
load_vscode_styles()

Notebook Conversion

Convert notebooks to different formats or clean them up:

import nbqol

# Convert notebook to VSCode-compatible Python script
nbqol.convert_to_vscode_script('my_notebook.ipynb', 'my_script.py')

# Clean a notebook by removing intermediates and large outputs
nbqol.clean_notebook('my_notebook.ipynb', remove_intermediates=True, remove_large_outputs=True)

# Convert notebook with customized options
nbqol.convert_notebook('my_notebook.ipynb', to_format='html',
                      clean_options={'remove_markdown': True})

# Alternative: use the process module directly
from nbqol.process import reformat
reformat.convert_to_vscode_script('my_notebook.ipynb', 'my_script.py')

Notebook Analysis

Analyze, execute, and verify notebooks:

import nbqol

# Execute a notebook and save the result
nbqol.execute_notebook('my_notebook.ipynb', 'executed_notebook.ipynb')

# Check if a notebook runs without errors
success, message = nbqol.check_notebook_runs('my_notebook.ipynb')

# Verify that certain variables exist in the notebook
results = nbqol.verify_notebook_variables('my_notebook.ipynb', ['df', 'model'])

# Analyze cell dependencies in a notebook
dependencies = nbqol.notebook_cell_dependencies('my_notebook.ipynb')

# Compare two notebooks
differences = nbqol.compare_notebooks('notebook1.ipynb', 'notebook2.ipynb')

# Alternative: use the process module directly
from nbqol.process import analyze
analyze.execute_notebook('my_notebook.ipynb', 'executed_notebook.ipynb')

Output Capture

Capture and manage output streams:

import nbqol

# Capture output from code execution
with nbqol.capture_output() as captured:
    print("This will be captured")
    import logging
    logging.warning("This warning will also be captured")

print("Captured stdout:", captured.stdout)
print("Captured stderr:", captured.stderr)
print("Captured logs:", captured.logs)

# Alternative: use the outputs module directly
from nbqol.outputs import capture_output
with capture_output() as captured:
    print("Alternative capture method")