Key Concept

Filter DataFrames using boolean conditions (“masks”) to select specific rows.

Basic Filtering

Single Condition

# Create boolean mask
mask = df['total_bill'] > 30
 
# Apply mask
large_bills = df[mask]

Equivalent shorthand:

df[df['total_bill'] > 30]

How It Works:

1. df['total_bill'] > 30 creates boolean Series (True/False values)
2. DataFrame returns rows where mask is True

---

Multiple Conditions

AND (&) Operator

# Both conditions must be True
df[(df['total_bill'] > 30) & (df['sex']!='Male')]

OR (|) Operator

# Either condition can be True
df[(df['day'] == 'Sun') | (df['day'] == 'Sat')]

---

isin() Method

Check if values are in a list of options:

# Weekend filter
options = ['Sat','Sun']
df['day'].isin(options)
 
df[df['day'].isin(['Sat', 'Sun'])]
 
# Equivalent to:
df[(df['day'] == 'Sat') | (df['day'] == 'Sun')]

---

Why Parentheses Matter

Python operator precedence can cause unexpected behavior:

# ❌ Wrong - evaluates 'Male' & df first!
df[df['total_bill'] > 30 & df['sex'] == 'Male']
 
# ✅ Correct
df[(df['total_bill'] > 30) & (df['sex'] == 'Male')]

---

Filtering Process

1. Create Mask: Generate boolean Series from condition
2. Apply Mask: Use df[mask] to filter rows
3. Combine Masks: Use logical operators (&, |, ~)

---

Key Takeaways

1. Boolean Masking: Core pandas filtering technique
2. Operator Precedence: Always use parentheses
3. isin() Efficiency: Cleaner than multiple OR conditions
4. Readability: Break complex filters into multiple lines