Regular Expressions

What are Regular Expressions?

Regular expressions (regex) are patterns to find text. They're like super-powered search.

Instead of finding exact text like "cat", you can find:

Any 3-letter word
All phone numbers
Any email address

Basic Pattern Matching

code.py

import pandas as pd

df = pd.DataFrame({
    'Text': ['cat', 'car', 'cut', 'dog']
})

# Find words starting with 'c'
matches = df[df['Text'].str.contains('^c', regex=True)]
print(matches)

Output:

  Text
0  cat
1  car
2  cut

Common Patterns

Pattern	Meaning	Example
^	Start of text	^hello
$	End of text	world$
.	Any single character	c.t matches cat, cut
\d	Any digit (0-9)	\d\d\d matches 123
\w	Any letter or number	\w+ matches hello
\s	Any space	hello\sworld
+	One or more	\d+ matches 1, 12, 123
*	Zero or more	ab* matches a, ab, abb

Find Numbers

code.py

df = pd.DataFrame({
    'Text': ['Order 123', 'Item 456', 'No number here']
})

# Extract numbers
df['Numbers'] = df['Text'].str.extract('(\d+)')
print(df)

Output:

             Text Numbers
0       Order 123     123
1        Item 456     456
2  No number here     NaN

Find Email Addresses

code.py

df = pd.DataFrame({
    'Text': ['Contact: john@gmail.com', 'Email sarah@yahoo.com', 'No email']
})

# Extract emails
df['Email'] = df['Text'].str.extract('(\w+@\w+\.\w+)')
print(df)

Output:

                      Text           Email
0    Contact: john@gmail.com  john@gmail.com
1    Email sarah@yahoo.com  sarah@yahoo.com
2                 No email             NaN

Find Phone Numbers

code.py

df = pd.DataFrame({
    'Text': ['Call 123-456-7890', 'Phone: 987-654-3210']
})

# Extract phone numbers
df['Phone'] = df['Text'].str.extract('(\d{3}-\d{3}-\d{4})')
print(df)

\d{3} means exactly 3 digits.

Replace Using Patterns

code.py

df = pd.DataFrame({
    'Phone': ['123-456-7890', '987-654-3210']
})

# Remove all non-digits
df['Clean'] = df['Phone'].str.replace('\D', '', regex=True)
print(df)

Output:

          Phone       Clean
0  123-456-7890  1234567890
1  987-654-3210  9876543210

\D means "not a digit"

Check Pattern Exists

code.py

df = pd.DataFrame({
    'Text': ['abc123', 'hello', '456def']
})

# Check if contains any digit
df['Has_Number'] = df['Text'].str.contains('\d', regex=True)
print(df)

Useful Patterns Cheat Sheet

What to Find	Pattern
Any number	\d+
Any word	\w+
Email	\w+@\w+.\w+
Phone (US)	\d{3}-\d{3}-\d{4}
Start with letter	^[A-Za-z]
End with number	\d$

Key Points

Regex finds patterns, not exact text
Use str.extract() to pull out matches
Use str.contains() to check if pattern exists
Use str.replace() with regex to clean data
Patterns look complex but follow simple rules

What's Next?

Learn to work with dates and times in your data.