python-intro-notes.txt

Leetcode Q https://leetcode.com/problems/climbing-stairs/
  recursion, dynamic programming
Big project? Sudoku solver?
  python flask backend rest-api front-end
Python environment:
  CPython, PyPy (not C, 8x faster), Jython (Java), IronPython (Dot Net),
  Anaconda (conda, math ML and AI), ActivePython,
  MicroPython, PyScript
  IDE: VSCode, PyCharm, Spyder, Komodo, Emacs, Vim
  modules:
    module.py, import module, sys.path, dir(module)
    if name == main
  pip list install
  venv > virtualenv. docker.
    python3 -m venv envs/myproject1
    source envs/myproject1/bin/activate
    deactivate

  

f- functions and generators (yield instead of return) 1. efficiency for long lists, 2. pausing complex calculations (co-routines)
- list comprehensions [x for x], generator comprehensions (x for x)

---
Lions Tigers and Bears, which data structure to use?
Tuples, Lists, Named Tuple, Dict, DefaultDict, Dataclass, Class
https://docs.python.org/3/
https://docs.python.org/3/library/index.html

primitive scalar data types (all immutable):
- int 5 (+ - * / // % etc) (int, long, bignum all handled invisibly in the background)
- float 5.2 (any scale, but approx 15 digits of accuracy)
- complex 5.2+3j
- bool (None, False, 0, '', [], (), {}) or (True, anything else) (and or not == != if b , comparisons < > <= >= is is not)
- str 'hello', "hello", """hello""" (+ is concatenate)
- immutable? you can set x=x+5 or x="hello "+x to set x to a newly created value, but you can't change x in place. The sequence type tuple is also immutable. Mutable x={'name':'peter','age':10}  x['age']=11. This means repeatedly concatenating more and more values onto the end of a tuple or string will be SLOW. Instead make a list and use join at the end to build it once, or use an f'string'

Sequences and Containers:
- tuple (1,2,"hello"), l[index], slices l[index:index], l*n, l1+l2, len(l), max(l), min(l), l.index(value), l.count(value), l.reverse(), l.sort()
	- function calls with dunder methods vs. class methods: len(s) vs s.len()  def \_\_len\_\_(): def len()
- list [1,2,"hello"]
- range range(10) range(1,10) range(2,10,2)
- dict { 'name':'sue','age':20} (defaultDict) d={'key':value,'key2':value2}, d[key], key in d, k not in d, for k in d, list(d), d.values(), del d[k], d.get(key, default)
- set set(), {'a','b','c'}, set( (1,2,3)) (frozenSet) value in set, len(set), for value in set, s.add(value), s.remove(value) s.discard(value), supersets and subsets with <= >=, union with S1|S2, intersection with s1&s2, difference with s1-s2
- stacks: list.pop() list.append()

Not built-in:
- from decimal import Decimal, x=Decimal(5.23)
- from collections import defaultdict, d=defaultdict(0)
- deque: fast append and pop from both ends

using Dicts and Tuples you can make any complicated data structure necessary, but using indexes can be error prone, naming your data reduces errors and is self-documenting:
custom data types: namedTuple, attrs, pydantic, dataclass, class

- from collections import namedtuple,
	- Student = namedtuple('Student', ['name', 'age', 'DOB'], defaults=[18, '20000101'])
	- s = Student('Nandini', 19, '2541997'), s=Student(name='nandini', age=19, DOB='2541997')
	- s[0], s.name
-  Dicts ~~ JSON
```
from dataclasses import dataclass
@dataclass
class DataClassCard:
    rank: str = 'A'
    suit: str = 'Diamonds'
queen_of_hearts = DataClassCard('Q', 'Hearts')
queen_of_hearts.rank
if queen_of_hearts == DataClassCard('Q', 'Hearts')

# regular classes
class RegularCard:
    def __init__(self, rank, suit):
        self.rank = rank
        self.suit = suit

```

Pandas:
load csv, json, excel
dtypes, shape, describe, info
df['age'].max() .sum() .count()
select cols: df['col'], df[ ['col','col2']]
select rows: filter: x[ x['mark']>70 ]
df.loc( conditional, cols)
df.iloc[ r1:r2, c1:c2] = x
join two tables newt = pd.merge( t1, t2, left_on="Name", right_on="name")
    how='inner' 'left' 'right' 'outer'
import matplotlib.pyplot as plt
df['newcol']= xxx