str.extract()

noun

say it: “S-T-R dot ek-STRAKT” /ˌɛs ti ˈɑr dɑt ɛkˈstrækt/

A pandas tool that pulls the part of each text value that matches a pattern, such as the number in "about 30 min", and marks values with no match as missing.

What does str.extract() do in pandas?

str.extract() keeps the part of each value that a regular expression's parentheses mark, such as the digits in "about 30 min". A value with no match becomes NaN. It suits a number written among words that vary, where str.replace() would need a call for every wording.

import pandas as pd
goals = pd.Series(["about 30 min", "28", "no goal"])
print(goals.str.extract(r"(\d+)"))   # 30, 28, NaN

In a sentence

  • “Rows where str.extract() found no digits came back as NaN.”
  • “str.extract() pulled the minutes out of goal times typed a dozen different ways.”

Level: Freelancer · More Python terms at this level

Updated