Python Basics II: Control Statements, Functions, Lambda & Packages

1. Review: return values (Week 2 Q7)

本講係一份 Jupyter notebook。開頭先講 Week 3 起嘅交功課方法：用 cell-level copy and paste 將題目相關嘅 cells 抄去新 notebook，以姓名 + student ID 命名，交 PDF 去 Moodle；唔好交成份有晒教學內容嘅原 notebook。

然後回顧 Week 2 Q7：

my_list = [1, 2, 3]
print(my_list.pop(0))      # 1
print(my_list.append(0))   # None

help(my_list.append)
# append(object, /) method of builtins.list instance
#     Append object to the end of the list.
help(my_list.pop)
# pop(index=-1, /) method of builtins.list instance
#     Remove and return item at index (default last).
#     Raises IndexError if list is empty or index is out of range.
my_list.pop.__doc__        # 直接攞 docstring 字串

2. Control statements: indentation & if

2.0 Groups of statements — indentation matters!

Python 用 colon (冒號) : 開始一組 statements，組內所有 statements 要用相同 indentation (縮進距離)。Python 冇規定一定要幾多個 spaces，但建議 4 spaces (PEP 8)。

2.1 if statement

• 用 if keyword；if 入面嘅 statements 要 indent 並且 indentation 一致，condition 為 True 時先執行。
• Condition 可以係 assertions (==, >, <…) 或者 function 返回嘅 Boolean value。
• 多過一個 condition 用 elif。

weather = 'Sunny'
print(f"The weather is {weather}.")
if weather == 'Sunny':
    print("Don't forget your suncream!")
elif weather == 'rainy':
    print("Dont forget your umbrella!")
else:
    print("Have a nice day!")
# The weather is Sunny.
# Don't forget your suncream!

Indentation problems (Lab Q1)

# Example 1 (錯)：if 後面一定要有 indented 嘅 action
if weather == 'sunny':
print('let"s go to the beach')        # IndentationError: expected an indented block

# 修正
if weather == 'sunny':
    print("let's go to the beach")    # 用雙引號包住，入面先可以用 apostrophe '

# Example 2 (錯)：同一 block 嘅 indentation 唔一致 / else 對錯位
# 修正
if weather == 'sunny':
    print("let's go to the beach")
    print('we will be happy')
else:
    print("let's stay at home")

elif vs else

• elif 要有 condition — 「if something occurs」。
• else 冇 condition — 「for all other cases」。else: 之後唔可以再有 branch，否則 SyntaxError: invalid syntax。

if weather == 'sunny':
    print('beach')
else:
    print('home')
elif weather == 'rainy':      # ✗ SyntaxError：else 一定係最後一個 branch
    print('home')

Lab Q2 參考答案

for i in [-5, 0, 3, 8]:
    if i < 0:
        print('Negative')
    elif i == 0:
        print('Equal to zero')
    elif i < 5:                 # 去到呢度已知 i > 0
        print('Positive but smaller than 5')
    else:
        print('Positive and larger than or equal to 5')
# Negative / Equal to zero / Positive but smaller than 5 / Positive and larger than or equal to 5

Conditional expression (ternary operator)

一行寫簡單 if-else：value_if_true if condition else value_if_false。適合根據 condition assign 值，可以用喺 assignments、function arguments、list comprehensions 等。Tip：只用於簡單 condition，保持可讀性。

weather = 'sunny'
result = "Let's go to the beach" if weather == 'sunny' else "Let's stay at home"
print(result)     # Let's go to the beach

3. Loops: for, range, while

3.1 for loops

for loop 用嚟 iterate 一個 iterable object 嘅元素；list、tuple、甚至 dict 都係 iterable。

for item in object:
    statement(s)

l1 = [1, 2, 3, 4, 5]
for i in l1:
    print(i)          # 1 2 3 4 5 (每行一個)

Supplementary: Arithmetic & Logical operators

Logical operators 組合多個 conditions，返回 True/False：

• and：兩個都 True 先 True — (5 > 3) and (2 < 4) → True
• or：至少一個 True — (5 > 3) or (2 > 4) → True
• not：反轉 — not(5 > 3) → False

Example: odd / even (list 或 tuple 都得)

l1 = (1, 2, 3, 4, 5)     # tuple 同 list 一樣可以 loop
for i in l1:
    if i % 2 == 0:
        print(f"{i} is even")
    else:
        print(f"{i} is odd")
# 1 is odd / 2 is even / 3 is odd / 4 is even / 5 is odd

3.2 Looping over dictionaries

Dict 係 iterable，可以 loop 佢嘅 keys、values 或 items。Python 3.7 起 dict 保留 insertion order；之前版本係 unordered (可以 loop，但次序冇保證)。Notebook 用嘅係 Python 3.10。

d1 = {"name": "aileen", "marks": 100}
type(d1.items())    # dict_items
type(d1.keys())     # dict_keys
type(d1.values())   # dict_values

for pair in d1:            print(pair)   # name / marks          ← 直接 loop dict 得到 key
for pair in d1.items():    print(pair)   # ('name','aileen') / ('marks',100)  ← tuple(key, value)
for pair in d1.keys():     print(pair)   # name / marks
for pair in d1.values():   print(pair)   # aileen / 100

for d in d1.items():
    print(f"the 'key' is {d[0]} and its value is {d[1]}")
# the 'key' is name and its value is aileen
# the 'key' is marks and its value is 100

3.3 range()

range 係 built-in function，產生一串數字，常用嚟 loop 指定次數：range(stop)、range(start, stop)、range(start, stop, step)。同 slicing 一樣 stop 唔包含。產生嘅 sequence 係 iterable 但唔係 list (要 list(range(...)) 先變 list)；直接 print 只會見到 range(1, 10)。

print(range(1, 10, 1))        # range(1, 10)  ← 唔會印出數字
for ii in range(1, 10, 1):
    print(ii)                 # 1 … 9 (冇 10)
for i in range(2, 10, 3):
    print(i)                  # 2 5 8

sum = 0                       # (slide 用 sum 做名；實際應避免覆蓋 built-in sum())
for i in range(13, 174, 7):
    sum += i
print(sum)    # 2070
print(i)      # 167 ← loop 完 i 保留最後一個值 (13 + 7×22 = 167；174 唔包含)

3.4 while loops

Condition 為 True 就一直 loop。要自己 increment / decrement iterator，否則 loop 永遠唔會停 (infinite loop)。

num = 0
while (num < 10):
    num += 1          # 忘記呢句 → while loop never ends
    print(num)        # 1 … 10

# 緊接住再行：
while (num in range(0, 10)):   # 此時 num == 10，唔喺 range(0,10) 入面 → 一次都唔行
    num = num + 1
print(num)            # 10

Lab Q3: sum multiples of 3 or 5 (0 – 99999)

a = 3
print((a%3==0) | (a%2==0))   # True  ← 有括號，先比較再 bitwise OR
print(a%3==0 | a%2==0)       # False ← | 優先過 ==，變成 a%3 == (0 | a%2) == 0 → 0 == 1 == 0 → False
print(a%3==0 or a%2==0)      # True  ← 建議用 or (precedence 低過 ==)

# for loop
total = 0
for i in range(0, 100000):          # 0 至 99999 (inclusive)
    if i % 3 == 0 or i % 5 == 0:
        total += i
print(total)    # 2333316668

# while loop
total, i = 0, 0
while i <= 99999:
    if i % 3 == 0 or i % 5 == 0:
        total += i
    i += 1                          # 自己 increment
print(total)    # 2333316668

4. Functions (and Methods)

Functions (同 methods) 係完成特定任務嘅 self-contained code modules：通常「take in」data、處理、「return」結果 (return 係 optional)。寫好之後可以重複使用，亦可以喺其他 functions 入面 call。Methods 係屬於特定 class 嘅 light functions，要先有該 class 嘅 object，再用 obj.abc()。

4.1 Built-in functions

Python interpreter 內置、永遠可用：print(), min(), max(), len(), type() 等 (完整列表：docs.python.org/3/library/functions.html)。

len("hello")     # 5
len([1, 2, 3])   # 3
len(1)           # TypeError: object of type 'int' has no len()

4.2 Custom-defined functions

點解唔 copy-paste 同一段 code？ Copy-paste 較易出錯；要改嘅時候要去每一個位置改。定義 function 就只需改定義嗰一處，之後 call 名就得，code 更乾淨清晰。

Syntax：

• 用 keyword def
• 圓括號放 argument(s)；可以係 variable length def func(*var)
• 用 colon :
• Function body 用相同 indentation
• 可以有或冇 return value

def calculate_area(width, height=3, *vars):
    """Calculate rectangle area from width and height."""   # docstring
    area = width * height
    if len(vars) > 0:
        for each in vars:
            print(each)
    return area

calculate_area(3, 5, ['random', 'input', 0])   # 印 ['random', 'input', 0]；return 15
calculate_area(1)                              # height 用 default 3 → return 3

def my_sum(a, b):
    print(f'{a} + {b} is {a+b}')
    return a + b

my_sum(1, 2)              # 1 + 2 is 3
my_sum('Aileen', 'Hou')   # Aileen + Hou is AileenHou  ← + 對 str 係連接
my_sum(b=5, a=7)          # 7 + 5 is 12 ；Out: 12 (notebook 只顯示最後一句嘅 return)

def myFunc():             # 冇乜意義但語法正確
    return True

def my_sum(a=5, b=5):
    return a - b
    print("Execution completed.")   # 永遠唔會執行 (return 之後)

my_sum(1, 2)       # -1  positional：a=1, b=2
my_sum(b=1, a=7)   # 6   keyword
my_sum()           # 0   default 5 - 5

def print_arguments(*args):
    print(args)            # ('A', 'B', 'C') ← tuple
    for arg in args:
        print(arg)
print_arguments("A", "B", "C")

Variable scope

將 variable (例如 a) 傳入 function 時，對於 immutable types (例如 int)，function 收到嘅效果等同一個 copy；function 入面改 a 唔會影響外面嘅 a。

a = 3
def my_update(a):
    a = a + 1
    print(f'inside the fucntion a is {a}')
my_update(a)
print(f'outside the function a is {a}')
# inside the fucntion a is 4
# outside the function a is 3

Lab Q4 & Q5 參考答案

# Q4 (for loop)
def small_positive_number(seq):
    result = []
    for x in seq:
        if (x > 0) and (x < 5):
            result.append(x)
    return result                      # return，唔係 print

# Q4 (while loop)
def small_positive_number(seq):
    result, i = [], 0
    while i < len(seq):
        if 0 < seq[i] < 5:
            result.append(seq[i])
        i += 1
    return result

small_positive_number([1, 2, 0, 4, 6, 5, 2, 1])   # [1, 2, 4, 2, 1]

# Q5
def my_func_varargs(arg1, *vartuple):
    print(arg1)
    for v in vartuple:
        print(v)
    return 1 + len(vartuple)           # 總 argument 數量

my_func_varargs(1, 'a', 'b', 'c')      # 印 1 a b c；return 4

5. Recursion & Debugging

5.1 Recursion

Recursion：function call 自己，用嚟解決可以拆成較細、相似 subproblems 嘅問題。

• 一定要有 base case，防止 infinite recursion，令 recursive calls 最終終止。
• 某些問題用 recursion 更優雅易明，但對大 input 可能用更多 memory、更慢 (相比 iterative)。

例子：binary search — 喺已排序 list 搵 target：睇中間 mid，等於就 return；細過 target 就去右半邊搵；大過就去左半邊搵。

def binary_search(nums, target, left=None, right=None):
    if left is None:
        left = 0
    if right is None:
        right = len(nums) - 1
    if left >= right:            # ← base case (有 bug！)
        return -1
    mid = left + (right - left) // 2
    if nums[mid] == target:
        return mid
    elif nums[mid] < target:
        return binary_search(nums, target, mid + 1, right)
    else:
        return binary_search(nums, target, left, mid - 1)

my_list = [1, 2, 3, 4, 5]
binary_search(my_list, 3)   # 2 ✓
binary_search(my_list, 5)   # -1 ✗ 應該係 4

5.2 Debugging

Debugging = 找出並修正 errors / bugs。Python 常用工具：

• Print statements：喺不同位置 print variables 嘅值。
• Exception handling：try-except 捕捉 exceptions，幫助定位錯誤來源。
• Logging：用 logging module 記錄訊息同事件。
• Debuggers：IDE / command-line debugger，set breakpoints、逐步執行、inspect variables。

# 5.2.1 用 print debug
def binary_search(nums, target, left=None, right=None):
    if left is None: left = 0
    if right is None: right = len(nums) - 1
    print(f"left = {left}, right = {right} -> {nums[left:right + 1]}")
    if left >= right:
        print("break because left >= right")
        return -1
    mid = left + (right - left) // 2
    if nums[mid] == target:
        return mid
    elif nums[mid] < target:
        return binary_search(nums, target, mid + 1, right)
    else:
        return binary_search(nums, target, left, mid - 1)

binary_search(my_list, target=5)
# left = 0, right = 4 -> [1, 2, 3, 4, 5]
# left = 3, right = 4 -> [4, 5]
# left = 4, right = 4 -> [5]
# break because left >= right
# Out: -1

# 用 logging (效果一樣，但可以按 level 開關，唔使刪 print)
import logging
logging.basicConfig(level=logging.DEBUG)
# 將上面 print(...) 改成 logging.debug(...) → 輸出 DEBUG:root:left = 0, right = 4 -> ...

Debugger (IDE)：set breakpoint (例如喺計 left / right 嗰行)，暫停時檢查 left, right, mid, nums[mid]：

6. Anonymous (lambda) functions

先分清：statements 包括 expressions；expressions 一定會產生一個值。例：x = 1 係 statement (assignment)，x == 1 係 expression (產生 True/False)。

Lambda function：細小、anonymous (冇名) 嘅 function，可以有多個 parameters 但只有一個 expression，又叫 lambda expression。

lambda arguments: expression

(lambda x, y: x - y)(3, 5)     # -2   ← 直接喺後面加 (args) call
sum_up = lambda x, y: x + y    # sum_up 只係成個 lambda body 嘅 "nickname"
sum_up(35, 5)                  # 40
(lambda x, y: x + y)(35, 5)    # 40   冇 nickname 都可以，用 () 包住 body

def remove_1st(x):             # 等價 full function
    return x[1:]
remove_1st_lambda = lambda x: x[1:]
remove_1st([1, 2, 3, 4, 5])          # [2, 3, 4, 5]
remove_1st_lambda(" Lecture 03")     # 'Lecture 03'  ← 對 string 一樣 work (slicing)

Lab Q6

is_even = lambda n: n % 2 == 0
print(is_even(4))   # True
print(is_even(7))   # False

Embedded usage: lambda 作為另一個 function 嘅 input

老師唔建議單獨用 lambda；佢主要用喺另一個 function 入面 — 作為 input parameter 或 return value。最經典：list.sort(key=...)。

id_string = ['a3', 'b1', 'c5', 'd2']
id_string.sort()                       # ['a3', 'b1', 'c5', 'd2']  按成個 string (第一個字母) 排
id_string.sort(key=lambda x: x[1])     # ['b1', 'd2', 'a3', 'c5']  按第二個字元 (數字) 排

Lab Q7 參考答案

student_dict = {"b003": "John", "a005": "Robin", "a002": "Aileen", "b001": "Mary"}

def student_sort(d):
    key_list = list(d.keys())
    key_list.sort(key=lambda x: int(x[-3:]), reverse=True)   # 按數字部分 (最後 3 位) 降序
    sorted_values = []
    for k in key_list:                 # ['a005', 'b003', 'a002', 'b001']
        sorted_values.append(d[k])
    return sorted_values

student_sort(student_dict)   # ['Robin', 'John', 'Aileen', 'Mary']

Lambda 作為 function 嘅 return value

def my_func(adder):
    return lambda x: x + adder

my_func(5)        # <function my_func.<locals>.<lambda>(x)>  ← 返回嘅係一個 function
my_func(5)(3)     # 8   等於 (lambda x: x + 5)(3)

7. Python packages & modules

• Module：單一 .py 檔，包含 functions、classes、variables，可以 import 使用。
• Package：有結構嘅 modules 集合，將相關 modules 組合埋一齊。
• 用 import 載入：import math、import pandas as pd。
• Python standard library 已包含好多 modules；numpy、pandas 等要另外安裝。
• 用 modules / packages 組織 code 令程式更 readable、reusable、maintainable。

7.1 pip commands

pip 係 Python 嘅 package installer，用嚟 install、update、manage packages。可以喺 terminal / command prompt 用，或者喺 Jupyter 用 ! (執行 terminal command) — 但更建議用 %pip，因為佢保證裝喺 notebook 當前使用嘅 environment；!pip install 可能裝咗去另一個 environment。

7.2 Install vs Import

Lab Q8

檢查並安裝：pandas (tables / DataFrames 嘅 data analysis & manipulation)、numpy (fast numerical operations、arrays)、seaborn (基於 matplotlib 嘅 statistical visualization)、plotly (interactive plotting)、scikit-learn (machine learning / data mining)。

%pip show pandas          # 每個 package 一個 cell
%pip show numpy
%pip show seaborn
%pip show plotly
%pip show scikit-learn    # 注意：install 名係 scikit-learn，import 名係 sklearn

import pandas as pd
import numpy as np
import seaborn as sns
import plotly.express as px
import sklearn

How to submit：將題目同答案放去另一個 notebook → export html → print as PDF → 確保所有 code、文字、outputs 都顯示 → 交 Moodle。

自我測驗 Self-test

x = 10
def g(x):
    x = x * 2
    return x
y = g(x)
print(x, y)

一頁總結 Cheat sheet