• Available License keywords: https://autopilot-docs.readthedocs.io/en/latest/license_list.html
• List of python exceptions
• Write Python - Run Everywhere - Beeware
• Python execution model
  • code-blocks, scopes, execution frames and exceptions
• Mock standalone function when unittesting
  • https://docs.python.org/3/library/unittest.mock.html#where-to-patch
• Unittest alternatives: https://pypi.org/project/chai/
• Copying files
• Doctest primer
• Logging module resources:
  • http://nathanielobrown.com/blog/posts/quick_and_dirty_python_logging_lesson.html
  • https://docs.python-guide.org/writing/logging/
• setup.cfg
  • example
• arrow - one class to manage dates
• requirements.txt vs install_requires
• pyenv - Alternate between different python versions
• Conflict resolution with argparse: https://docs.python.org/2.4/lib/optparse-conflicts-between-options.html

    parser = OptionParser(conflict_handler="resolve")
    parser.add_option("-n", "--dry-run", ..., help="do no harm")
    parser.add_option("-n", "--noisy", ..., help="be noisy")
  

• python performance hints
• concerns when packaging c extensions in python packages
• dynamic-programming-in-python
• memory-tempfile
  • When to use /tmp and when to use /dev/shm: https://superuser.com/a/45509/369517
  • More information: https://superuser.com/a/1030777/369517
• Convert time since epoch to datetime:
  • datetime.datetime.fromtimestamp(1284286794)
• Python staticmethod vs classmethod:
  • https://stackoverflow.com/a/1669524/2843583
  • staticmethods are basically freestanding functions grouped under the class scope; classmethods are still methods of a class.
• FactoryBoy: Better fixtures for your Django tests
• Is Python an interpreted or a compiled language?
  • Python is a language specification. One of it's implementations CPython, compiles the source code into bytecode, which then is run (interpreted) inside the Python VM.
  • Link: https://stackoverflow.com/a/2998228/2843583
• Python is caching the first 257 integers in memory:

    # Same identity all the time
    In [49]: id(256)
    Out[49]: 9796768
  
    # Generated on the fly
    In [50]: id(257)
    Out[50]: 139740392044752
  

• bdist: build distribution
• sdist: source distribution -> tar.gz
• pip : pip installs packages

• https://github.com/thclark/rstcloth
  • Commit hooks for running autoformatting
  • Based on a cookiecutter template

>>> cli_timestamp_to_ns("123") # doctest: +IGNORE_EXCEPTION_DETAIL
Traceback (most recent call last):
python_utils.custom_exceptions.InvalidCLITimestampError: Invalid timestamp provided: 123

• Run a single test method (or a single test class or ...): python3 -m unittest package.path.to.class.or.method

• See loaded modules in python: set(sys.modules) & set(globals())
• Pip downloads every single version of a package:
  • Use --use-deprecated=legacy-resolver instead - https://github.com/pypa/pip/issues/8713

Print numpy array without truncation:

np.set_printoptions(threshold=sys.maxsize)

Quaternions:

• Form: w, x, y, z

• Preferred extension: pkl / pickle - https://stackoverflow.com/questions/40433474/preferred-or-most-common-file-extension-for-pickle-files/40433504
• Use the dump, load methods - open file inline for auto-close

• Python3 pathlib library: https://realpython.com/python-pathlib/

There's a gazzilion of 2D statistics/plotting libraries in python. Just select a single damn one and stick to it - None of it supports 3D visualisation

• bokeh: https://bokeh.pydata.org/en/latest/docs/user_guide/quickstart.html
• gpgpy
• plotly
• seaborn
• altair
• pygal

Only thing that's available (~end 2019) is matplotlib. Just use it...

• https://medium.com/@sebastiannorena/3d-plotting-in-python-b0dc1c2e5e38
• https://matplotlib.org/mpl_toolkits/mplot3d/tutorial.html
• https://plot.ly/python/3d-surface-plots/
• Matplotlib rcParams available parameters: https://matplotlib.org/3.1.0/tutorials/introductory/customizing.html

parser = argparse.ArgumentParser(description="Help line 1\n" +
                                             "Help line 2",
                                 formatter_class=lambda prog: argparse.RawTextHelpFormatter(prog, width=99999))

Example usage: https://www.data-structures-in-practice.com/hash-tables/

Use the dis standard library module: https://docs.python.org/3/library/dis.html Use it e.g., on a function:

import dis
f = lambda x: x + 2
dis.dis(f)
1           0 LOAD_FAST                0 (x)
            2 LOAD_CONST               1 (2)
            4 BINARY_ADD
            6 RETURN_VALUE

Normally you'd do this as part of your bashrc/fish.config file. But if you haven't, here's the commands to do it in the current shell instance

fish )
    set -x PATH \"${PYENV_ROOT}/bin\" \$PATH
    status --is-interactive; and . (pyenv init -|psub)
    status --is-interactive; and . (pyenv virtualenv-init -|psub)
    ;;
* )
    PYENV_ROOT="$HOME/.pyenv"
    export PATH="${PYENV_ROOT}/bin:$PATH"
    eval "$(pyenv init -)"
    eval "$(pyenv virtualenv-init -)"

• Compile python versions from source:
  • You'll need to install OpenSSL first: apt-get install libssl-dev libffi-dev
• Format all .py files with black autoformatter:
  • fd -e py | xargs -I{} black --line-length 95 -t py36 {}
• Render a html python string in the browser:

  # I assume that the text was acquired using the requests library
  import webbrowser
  
  with open("a.html", "w") as f:
      f.write(resp.text)
  
  webbrowser.open("a.html")
  

from sympy import var, cos, plot
x = var('x')
p = plot(cos(x), line_color='r')
p = plot(cos(2*x), line_color='g')
p = plot(cos(4*x), line_color='b')
pl.extend(plot(cos(2*x), line_color='g'))
pl.show()

numpy.set_printoptions(linewidth=160)

(Pdb++) p T
array([[ 1.00000000e+00,  1.00284240e-18, -2.28654607e-18,
         0.00000000e+00],
       [ 1.13765545e-18,  1.00000000e+00,  1.29264105e-19,
         0.00000000e+00],
       [-1.39240192e-17,  1.34405256e-19,  1.00000000e+00,
         3.46944695e-18],
       [ 0.00000000e+00,  0.00000000e+00,  0.00000000e+00,
         1.00000000e+00]])
(Pdb++) numpy.set_printoptions(linewidth=160)
(Pdb++) p T
array([[ 1.00000000e+00,  1.00284240e-18, -2.28654607e-18,  0.00000000e+00],
       [ 1.13765545e-18,  1.00000000e+00,  1.29264105e-19,  0.00000000e+00],
       [-1.39240192e-17,  1.34405256e-19,  1.00000000e+00,  3.46944695e-18],
       [ 0.00000000e+00,  0.00000000e+00,  0.00000000e+00,  1.00000000e+00]])
(Pdb++)

https://github.com/albertlauncher/python/pull/96/files

bus = dbus.SessionBus()
notify = dbus.Interface(bus.get_object(dbusItem, dbusPath), dbusInterface)
notify.Notify(__prettyname__, 0, iconPath, title, text, '', '', 0)

spec = importlib.util.spec_from_file_location("bing", "plugins/image_search/bing.py")
foo = importlib.util.module_from_spec(spec)
spec.loader.exec_module(foo)

cl = foo.MyClass()

• Rename just a single column in pandas: df.rename(columns = {'old_name':'new_name'}, inplace = True)
• Drop column by index i: df.drop(df.columns[i], axis=1)
• Just plain nice: df['Change (%)'] = df['Change (%)'].str.strip('%').astype(float)

• https://github.com/myint/scspell
  • See colcon-zsh (or any other colcon plugin) as an example

Every module and object in python has a __dict__ attribute (you can think of all the modules as dictionaries that contain their variables, and functions)

• Except int - doesn't have __dict__

Use self.__class__.__bases__

Link: https://stackoverflow.com/a/10092179/2843583

• Using Pip
• Debugging: Keyring Issue
• Debugging: unpackb-msgpack Error
• Simulate pushd using @contextmanager
• Poetry Package Manager
• Handling Integer Objects
• Debugging Tools
• VLC Python API - Read metadata from network stream
• Iterators vs Generators
• Concerns Packaging C Extensions in Python Packages
• Python Object Unique Name
• Check if PID exists
• Thought on Python coding standards
• Concurrency and AsyncIO
• functools
• decorators
• decorator once per minute
• psrecord
• Call subprocess and print its output as it comes / live
• Avoid if-initialisation cases
• unittest runner script
• Find System Installed Version That Matches
• Logging
• Code Coverage
• Decorator with Arguments
• ipython Interpeter
• Choosing TOML Over Other Formats
• Importing in Python
• Readline: Enable TAB Completion
• Click - List all visible, registered subcommands
• pdb / ipdb
• Upload to PyPI
• Decorators for Class and Object Attributes
• Mocking in Python
• mypy - annotate self class itself
• Find N-th Occurrence
• Pytest Testing Suite
• Mypy Cheatsheet
• Invoke a click subcommand - Have all subcommands share a list of arguments
• Regexp: Match URLs
• Packaging Options
• numba
• Bumping the version of your Python Project
• Implementations of the Python Specification
• pyproject.toml
• simple decorator
• pyc-files