Skip to content

Latest commit

 

History

History
185 lines (164 loc) · 4.21 KB

topics.md

File metadata and controls

185 lines (164 loc) · 4.21 KB

Topics to do for Python presentations

  • pygdb gdb for python

    • debugging python not in ide

  • eval() and exec() -> run python code from a string

    • modifiable code

  • number types -> imaginary, Decimal(), Fraction()

  • string types:

    • string (unicode)
    • bytes (ascii)
    • r, f other prefixes for string

  • string encoding decoding

  • '...' dot dot dot syntax

  • functional programming

    • lambda functions
    • list, map, filter, accumulate, chain, reverse, sort, sorted, reduce
    • first-class functions
      • objects can be functions, functions can be passed, functions can return functions, etc
    • closure, nonlocal
    • partial application of functions
      • from functools import partial
    • currying -> f(1)(2)(3)
    • last statement in function implicit return

  • walrus operator

  • variable scope differences

    • can access variable outside of an if, for, while, try
    • can't access outside of a function
    • as opposed to curly braces languages that don't allow this
    • use del to delete variables

  • del syntax

    • use to delete variables
    • delete items in list, dictionary?
    • what else?

  • Counter()

  • OrderedDict()

  • collections module

    • deque

  • Enum

    • unique, auto values

  • dunder methods

    • __str__
    • __repr__
    • eq, ge, lt
    • getattribute, setattr, getattr

  • logging

  • context managers

  • coroutines

  • python project management

    • __init__.py
    • requirements.txt

  • python 2 to 3 conversion

  • testing

    • unittest
    • pytest

  • simple standard library stuff:

    • calendar
    • os
    • os.path
    • shutil
    • urllib.request
    • re
    • timefile
    • zipfile
    • threading
    • multuprocessing
    • datetime
    • textwrap
    • html.parser
    • json
    • itertools.permutations

  • useful libraries:

    • requests
    • tkinter
    • numpy
    • simpleaudio
    • opencv
    • imageio
    • matplotlib
    • flask
    • jinja2
    • csv
    • pillow
    • pandas
    • xml
    • lxml
    • BeautifulSoup
    • selenium
    • nltk
    • sqlite3
    • redis

  • FASTER python

    • perf_counter()
    • timeit module
    • line_profiler
    • tracemalloc
    • look at python disassembly to optimize
    • attributes stored in dictionary, dictionary only 1/3 full waste a lot of memory
      • use __slots__ to reduce memory
    • lru_cache from functools import lru_cache
    • Joblib
    • parallel computing
      • Amdahl's law
      • io bound -> threads
        • ThreadPoolExecutor
        • threads as classes
        • thread as function
        • locking
      • cpu bound
        • ProcessPoolExecutor
      • asyncio
      • multi thread
      • multi processor?
      • multi python interpreters?
      • GIL problem
    • numpy
    • numba
    • cython
    • pypy
    • c extensions
    • strace
    • bloom filter
    • python async, await, async with, async for ... in ...

  • pyenv - separate package environments

Topics in process of creating presentation for:

Presentation 4:

  • generators
    • last vs. non-lazy computation
    • generator expressions
    • generator functions with yield
      • recursive generators with yield from
    • speed / size comparison
    • infinite data structures with generators
  • need something else

Topics done in presentations before

Presentation 1: (Oct 16, 2020)

  • *args and **kwargs
  • defaultdict
  • comprehension -> list, dict, set comprehension
  • faster python:
    • access global variables is SLOWER than local variables
    • accessing functions through obj.func is slower than storing func as local variable (cost for function lookup on object)
  • what's new in python 3.9

Presentation 2: (planned Oct 30th, 2020)

  • OOP stuff (kind of)
    • private attributes, hidden attributes, name mangling
    • Abstract classes from abc import ABC, abstractmethod
    • multiple inheritance and name conflicts, __mro__
    • interfaces through stateless abstract base classes and multiple inheritance
    • data class -> easier init functions (kind of like namedtuple)
      • immutable data class

Presentation 3:

  • string formatting
    • string append
    • %
    • format
    • f-strings
    • Template strings
  • decorators
    • manually
    • simple
    • with decorator arguments
    • with function arguments
    • decorators as Classes (have internal state)
    • decorators as cache / memoization