Label-based row inspection
You need to inspect or transform each row using column labels and the row's index.
Pandas · Module
df.iterrows
Iterate over DataFrame rows as index-and-Series pairs. Each row is exposed as a Series, so its values may not retain the original column dtypes; itertuples() is generally faster and preserves types better.
How it works
Call iterrows() to obtain an iterator whose items contain a row's index label and the row data as a Series. Because each row is represented as a Series, values can be coerced to a common dtype; use itertuples() instead when preserving types and improving iteration speed matter.
Syntax
DataFrame.iterrows()Returns
An iterator yielding pairs of the row index label (or a tuple of labels for a MultiIndex) and the row data as a Series.
Key points
- The method provides row-wise iteration as pairs containing an index and a Series.
- Each yielded index item identifies the corresponding DataFrame row.
- With a MultiIndex, the yielded row identifier is a tuple of index labels.
- The second item in each pair is the row's data represented as a Series.
- For type-preserving row access,
itertuples()returns namedtuples containing the row values. - The documentation identifies
itertuples()as generally faster than this method.
Common mistakes
- Do not rely on changing the yielded Series to update the original DataFrame; the iterator may provide a copy rather than a view.
- Mutating the row object during iteration may not modify the DataFrame.
Example: Iterate over DataFrame rows
import pandas as pd
df = pd.DataFrame({"name": ["Ada", "Grace"], "score": [10, 12]})
for index, row in df.iterrows():
print(index, row["name"], row["score"])
Iterate through rows while accessing each row by column label.
When to use it
Simple row processing
A small or moderate DataFrame requires straightforward row-wise logic and dtype coercion is acceptable.
Index-aware processing
You need to handle each record together with its index label.
Interview questions
QWhat does DataFrame.iterrows() yield?
Each iteration yields a two-item tuple: the row's index label and a Series containing that row's values. For a MultiIndex, the index item is a tuple of labels.
Follow-up: How would you iterate while retaining column dtypes? Use itertuples(), which yields namedtuples and is generally faster.
QDoes iterrows() preserve dtypes?
No. The row Series can promote values to a shared dtype, so the original column dtypes are not necessarily retained in each row. This differs from DataFrame column-wise dtype preservation.
Follow-up: What behavior demonstrates this? An integer value in a row containing a floating-point value can be represented as a float in the yielded Series.
QWhen should iterrows() be avoided?
Use iterrows() when row-wise Series access is convenient and the performance and dtype limitations are acceptable. Avoid treating the yielded objects as guaranteed writable views of the DataFrame.
Follow-up: What is the preferred alternative for faster, type-consistent row iteration? Use itertuples().
Related
DataFrame.itertuplesDataFrame.items