เมธอด query()

ทำไมจึงควรใช้ query()

import pandas as pd

df = pd.DataFrame({
    'name': ['Alice', 'Bob', 'Carol', 'Dave'],
    'age': [25, 35, 28, 45],
    'salary': [50000, 90000, 70000, 120000]
})

# Boolean indexing style
result1 = df[(df['age'] > 27) & (df['salary'] > 60000)]

# Equivalent query() style
result2 = df.query('age > 27 and salary > 60000')

print(result2)
#     name  age  salary
# 1    Bob   35   90000
# 2  Carol   28   70000

กฎไวยากรณ์ภายในสตริง Query

import pandas as pd

df = pd.DataFrame({
    'product': ['A', 'B', 'C', 'D'],
    'price': [10, 200, 50, 300],
    'in_stock': [True, False, True, True]
})

# Use 'and', 'or', 'not' in query strings
result = df.query('price > 40 and in_stock == True')
print(result)
#   product  price  in_stock
# 2       C     50      True
# 3       D    300      True

การอ้างอิงตัวแปร Python ด้วย @

import pandas as pd

df = pd.DataFrame({
    'city': ['NYC', 'LA', 'Chicago', 'Houston'],
    'population': [8336817, 3979576, 2693976, 2320268]
})

min_pop = 3_000_000  # Python variable

# Use @ to inject the variable into the query
result = df.query('population > @min_pop')
print(result)
#    city  population
# 0   NYC     8336817
# 1    LA     3979576

การต่อคำเรียก query()

import pandas as pd

df = pd.DataFrame({
    'region': ['North', 'South', 'North', 'East', 'South'],
    'year': [2022, 2022, 2023, 2023, 2023],
    'revenue': [100, 200, 150, 80, 300]
})

# Chain two query calls
result = (df
    .query('year == 2023')
    .query('revenue > 100')
    .sort_values('revenue', ascending=False)
)
print(result)
#    region  year  revenue
# 4   South  2023      300
# 2   North  2023      150

เปรียบเทียบ query() กับการทำดัชนีแบบบูลีน

import pandas as pd

df = pd.DataFrame({
    'A': range(10),
    'B': range(10, 20)
})

# Boolean indexing
result_bool = df[(df['A'] > 2) & (df['B'] < 18) & (df['A'] != 5)]

# Equivalent query — much more readable
result_query = df.query('A > 2 and B < 18 and A != 5')

print(result_query)
#    A   B
# 3  3  13
# 4  4  14
# 6  6  16
# 7  7  17

การเปรียบเทียบสตริงใน query()

import pandas as pd

df = pd.DataFrame({
    'country': ['USA', 'UK', 'Canada', 'USA', 'Germany'],
    'sales': [400, 150, 200, 600, 300]
})

# Single quotes for the string value inside double-quoted query
result = df.query('country == "USA"')
print(result)
#   country  sales
# 0     USA    400
# 3     USA    600

การใช้ตัวดำเนินการ in และ not in

import pandas as pd

df = pd.DataFrame({
    'status': ['active', 'inactive', 'pending', 'active', 'closed'],
    'amount': [100, 200, 50, 300, 150]
})

# Keep rows where status is in a specific list
result = df.query('status in ["active", "pending"]')
print(result)
#     status  amount
# 0   active     100
# 2  pending      50
# 3   active     300

# Exclude certain statuses
excluded = df.query('status not in ["closed", "inactive"]')
print(excluded.shape)  # (3, 2)

ตัวกรองช่วงตัวเลขด้วย query()

import pandas as pd

df = pd.DataFrame({
    'product': ['A', 'B', 'C', 'D', 'E'],
    'price': [5, 50, 150, 300, 500]
})

# Chained comparison — only works inside query()
result = df.query('50 <= price <= 300')
print(result)
#   product  price
# 1       B     50
# 2       C    150
# 3       D    300

ข้อจำกัดของ query()

import pandas as pd

df = pd.DataFrame({'first name': ['Alice', 'Bob'], 'class': [1, 2]})

# Backtick-quote column names with spaces or reserved words
result = df.query('`first name` == "Alice" and `class` == 1')
print(result)
#   first name  class
# 0      Alice      1

ตัวอย่างเชิงปฏิบัติ: การกรองคำสั่งซื้อ

import pandas as pd

orders = pd.DataFrame({
    'region': ['East', 'West', 'East', 'West', 'East'],
    'year': [2023, 2023, 2024, 2024, 2024],
    'revenue': [100, 200, 300, 400, 500]
})

min_year = 2024

# Filter to recent orders in the East region, then sum revenue
summary = (
    orders
    .query('year >= @min_year and region == "East"')
    .groupby('region')['revenue'].sum()
)
print(summary)
# region
# East    800
# Name: revenue, dtype: int64

แนวทางปฏิบัติที่ดีในการต่อเมธอดด้วย query()

import pandas as pd

df = pd.DataFrame({
    'dept': ['Eng', 'HR', 'Eng', 'Sales', 'HR'],
    'salary': [90000, 50000, 120000, 70000, 55000],
    'years': [3, 5, 8, 2, 4]
})

result = (
    df
    .query('years > 2 and dept != "HR"')
    .assign(bonus=lambda x: x['salary'] * 0.1)
    .sort_values('salary', ascending=False)
)
print(result[['dept', 'salary', 'bonus']])
#    dept  salary   bonus
# 2   Eng  120000  12000.0
# 0   Eng   90000   9000.0

ตรวจสอบความเข้าใจอย่างรวดเร็ว

ทบทวนบทเรียน