Data Manipulation

Pandas Basics

1 min read

Published Sep 29 2025, updated Aug 17 2026

PandasPython

Renaming Columns

You can rename all columns at once or specific ones.

# Rename all columns by assigning a new listdf.columns = ["Name", "Age", "Salary"]# Rename selected columnsdf.rename(columns={"old_name": "new_name"}, inplace=True)

Use .rename() when you only want to change a few columns.

Selecting Columns by Data Type

.select_dtypes() helps filter numeric, categorical, or boolean columns.

# Select only numeric columnsdf_numeric = df.select_dtypes(include=["number"])# Select only object (string) columnsdf_object = df.select_dtypes(include=["object"])# Exclude float columnsdf_no_float = df.select_dtypes(exclude=["float"])

Converting to NumPy

Access the underlying NumPy representation of the DataFrame.

# returns a 2D NumPy arraydf_array = df.values

This strips column and index labels (just raw data).

Replacing Data

.replace() can substitute values in the whole DataFrame or specific columns.

# Replace a single valuedf.replace(0, pd.NA, inplace=True)# Replace multiple valuesdf.replace([1,2,3], [10,20,30], inplace=True)# Replace in a single columndf["col"].replace("?", "Unknown", inplace=True)

Changing Data Types

.astype() is used to convert a column to a new data type.

# Convert a column to intdf["Age"] = df["Age"].astype(int)# Convert multiple columnsdf = df.astype({"Age": "int32", "Salary": "float"})

Mapping Data

.map() can apply a function, dictionary mapping, or Series to each element in that column.

df["col"].map({"M": "Male", "F": "Female"})

This replaces "M" with "Male" and "F" with "Female" in that column.

You can also pass a function:

df["col"].map(str.lower)

applymap() can apply a function elementwise to every single cell in the entire DataFrame.

df.applymap(str.upper)

Converts every value in the DataFrame to uppercase strings.