describe()

noun

say it: “dih-SKRYB” /dɪˈskraɪb/

A pandas method that gives a quick statistical summary of each number column, such as its average, smallest and largest values.

What does describe() do in pandas?

describe() sums up each number column of a DataFrame in eight rows: count, mean, standard deviation (std), min, the three quartiles (25%, 50% and 75%) and max. A min or max that can't be right, such as a price of -4, shows up here first.

import pandas as pd
sales = pd.read_csv("sales.csv")
sales.describe().round(2)    # count, mean, std, min, 25%, 50%, 75%, max for units and price

In a sentence

  • “She ran describe() to see the typical number of units sold.”
  • “describe() showed a smallest price of 0, which nobody charges.”

Level: Journeyman · More Python terms at this level

Updated