shape is an attribute, not a method — you access it directly without parentheses, the same convention as a NumPy array's .shape, since a DataFrame is fundamentally a 2D structure under the hood. It's the fastest way to check a DataFrame's size, and is commonly used to sanity-check the result of an operation, like verifying a merge or filter produced the expected number of rows.

1Understanding df.shape

💡

Remember df.shape has no parentheses — it's an attribute, not a method call, the same as a NumPy array's .shape; writing it as a function call raises a TypeError since a tuple isn't callable.

editor.html

import pandas as pd

df = pd.DataFrame({"a": [1, 2, 3], "b": [4, 5, 6]})
print(df.shape)

localhost:3000

2Practical Example

Here is a real-world application of df.shape showing how it is used in production Pandas code.

editor.html

import pandas as pd

df = pd.DataFrame({"a": range(100), "b": range(100)})
filtered = df[df["a"] > 50]
print(f"Rows: {filtered.shape[0]}, Columns: {filtered.shape[1]}")

localhost:3000

3Best Practices

Follow these guidelines when working with df.shape:

1. Check df.shape immediately after operations like merging, filtering, or dropping rows/columns, to quickly verify the result matches expectations

2. Access shape[0] for row count or shape[1] for column count directly, instead of calling len(df) and len(df.columns) separately

3. Remember shape is an attribute, no parentheses, unlike most DataFrame inspection methods like head() or describe()

⚠️

Tip: Remember df.shape has no parentheses — it's an attribute, not a method call, the same as a NumPy array's .shape; writing it as a function call raises a TypeError since a tuple isn't callable.

editor.html

import pandas as pd

df = pd.DataFrame({"a": [1, 2, 3], "b": [4, 5, 6]})
print(df.shape)

localhost:3000

Examples

Example 01Basic Usage

import pandas as pd

df = pd.DataFrame({"a": [1, 2, 3], "b": [4, 5, 6]})
print(df.shape)

Example 02Advanced Example

import pandas as pd

df = pd.DataFrame({"a": range(100), "b": range(100)})
filtered = df[df["a"] > 50]
print(f"Rows: {filtered.shape[0]}, Columns: {filtered.shape[1]}")

Best Practices

Check df.shape immediately after operations like merging, filtering, or dropping rows/columns, to quickly verify the result matches expectations
Access shape[0] for row count or shape[1] for column count directly, instead of calling len(df) and len(df.columns) separately
Remember shape is an attribute, no parentheses, unlike most DataFrame inspection methods like head() or describe()

Interview Question

Why does df.shape[0] give the number of rows rather than the number of columns?▼

Hint: Think about how shape's tuple ordering matches NumPy's array convention.

shape follows the same convention as a NumPy array's shape tuple, where each position corresponds to a dimension, and for a DataFrame, the first dimension, axis 0, is the row axis, and the second, axis 1, is the column axis. So shape[0] is the size along the row axis, the row count, and shape[1] is the size along the column axis, the column count — consistent with how axis numbering works throughout pandas and NumPy more broadly, like the axis argument in functions such as sum() or mean().

Exercises

MediumPractice using df.shape in a real scenario.

View Solution

import pandas as pd

df = pd.DataFrame({"a": [1, 2, 3], "b": [4, 5, 6]})
print(df.shape)

Frequently Asked Questions

Why does df.shape[0] give the number of rows rather than the number of columns?▼

Related Functions

df-dtypesndarray-shapedf-columns