Label-based selection
Retrieve specific records or fields by their index and column labels.
Pandas · Const
df.loc
DataFrame.loc selects and assigns DataFrame rows and columns by label, with optional boolean-array filtering. Label slices include both endpoints, and numeric keys remain labels rather than positions.
How it works
DataFrame.loc is a label-oriented indexer for selecting or assigning rows and columns. It accepts individual labels, label collections, inclusive label slices, boolean masks, callable indexers, and alignable objects.
Syntax
df.loc[row_indexer, column_indexer]Parameters
| Name | Type | Description |
|---|---|---|
indexer | label, list-like, slice, boolean array, alignable Series/Index, or callable | Row and column indexers supplied inside df.loc[row_indexer, column_indexer]. Supported forms include labels, lists or arrays of labels, inclusive label slices, boolean arrays, alignable boolean Series, alignable Index objects, and callables. |
Returns
A cross-section consisting of selected row(s) or column(s) from the DataFrame; the exact result type depends on the indexers.
Key points
- Use loc to retrieve or assign groups of rows and columns using labels or boolean selection.
- Its main indexing model is labels, with boolean arrays also supported.
- Pass a list or array when selecting multiple, explicitly named labels.
- Use label slices to select a contiguous label range.
- A positional boolean mask must have the same length as the axis it filters.
- A boolean Series can filter rows while pandas aligns its index to the target.
- When an alignable Index is used, the result retains that Index as its selection index.
Common mistakes
- Label-based slices include both boundary labels, unlike ordinary Python slices.
- Numeric keys are still treated as index labels, not positional offsets; use iloc for integer-position access.
- An alignable boolean Series is matched to the DataFrame index before it filters rows.
Example: Select and update by labels
import pandas as pd
df = pd.DataFrame(
{"name": ["Ada", "Linus", "Grace"], "score": [95, 82, 91]},
index=["a", "b", "c"],
)
selected = df.loc[["a", "c"], ["name", "score"]]
high_scores = df.loc[df["score"] >= 90, "name"]
df.loc["b", "score"] = 88
print(selected)
print(high_scores)Uses label lists, a boolean condition, and a label pair to select and update DataFrame values.
When to use it
Boolean filtering
Keep rows satisfying a condition and optionally select particular columns.
Label-based assignment
Change selected cells or ranges without relying on positional offsets.
Inclusive label slicing
Select an inclusive range of ordered index or column labels.
Related
DataFrame.ilocDataFrame.atDataFrame.iatSeries.loc