Tutorials › Python › pandas Series

pandas Series

The one-dimensional pandas data structure — a single labelled column of data

A pandas Series is a one-dimensional labelled array — think of it as a single column in Excel with row labels. Every column in a pandas DataFrame is a Series. Understanding Series is the foundation for understanding DataFrames. When you select one column from a DataFrame (df["Sales"]), you get a Series back.

Example

Creating and working with Series
import pandas as pd

# Create a Series
sales = pd.Series([45000, 72000, 38000, 91000], name="Sales")
# 0    45000
# 1    72000
# 2    38000
# 3    91000
# Name: Sales, dtype: int64

# With custom index (like row labels)
sales = pd.Series(
    [45000, 72000, 38000, 91000],
    index=["Delhi", "Noida", "Gurgaon", "Mumbai"],
    name="Sales"
)
# Delhi      45000
# Noida      72000
# Gurgaon    38000
# Mumbai     91000

# Access values
sales["Noida"]      # 72000
sales[["Delhi", "Mumbai"]]  # two cities

# Statistics
sales.sum()    # 246000
sales.mean()   # 61500.0
sales.max()    # 91000
sales.idxmax() # "Mumbai"  — index of max value

Key Points

  • ✓Series = one column with row labels (index)
  • ✓df["column_name"] returns a Series; df[["col1","col2"]] returns a DataFrame
  • ✓Series.value_counts() — very useful: counts unique values
  • ✓Series.unique() — list of unique values; .nunique() — count of unique values
  • ✓Boolean Series: df["Salary"] > 60000 returns True/False for each row — used for filtering

Practice Question

What does series.value_counts() do?