Learn the key differences between loc and iloc in Pandas. Understand when to use label-based (loc) vs integer-based (iloc) indexing for efficient data manipulation.
iloc (Integer-Based Indexing) selects rows and columns using integer positions (like Python list indexing). It is position-based and ignores the actual index labels.
0 for the first row).[start:end] (exclusive of end).df.iloc[0] # First row (position 0)
df.iloc[1:4] # Rows 2, 3, and 4 (positions 1 to 3, end-exclusive)
df.iloc[:, 2] # All rows, third column (position 2)
loc (Label-Based Indexing) selects rows and columns using explicit labels (index names or boolean conditions). It is label-based and includes the end of slices.
df.loc[0:3] includes index 3).df.loc[0] # Row with index label `0`
df.loc[0:3] # Rows from label `0` to `3` (inclusive)
df.loc[df['col'] > 10] # Filter rows where 'col' > 10
df.loc[:, 'column_name'] # All rows, specific column by name
loc vs iloc – Key Comparisons*Pandas provides two primary methods for data selection: loc and iloc. While both are used for indexing, they work differently.
| Feature | loc |
iloc |
|---|---|---|
| Selection Type | Label-based | Integer position-based |
| Syntax | df.loc[row_label, column_label] |
df.iloc[row_idx, column_idx] |
| Includes Last Index? | Yes (inclusive) | No (exclusive, like Python slicing) |
| Works With | Row/column names (labels) | Row/column integer positions |
| Supports Boolean Masks? | Yes | Yes |
| Example | df.loc[2:5, 'Name'] |
df.iloc[2:5, 0] |
loc vs iloc?✔ Use loc when selecting data by labels (e.g., column names, index labels).
✔ Use iloc when selecting data by numeric positions (like list/array indexing).
import pandas as pd
data = {'Name': ['Alice', 'Bob', 'Charlie', 'David'],
'Age': [25, 30, 35, 40]}
df = pd.DataFrame(data, index=['a', 'b', 'c', 'd'])
# Using loc (label-based)
print(df.loc['a':'c', 'Name']) # Includes 'c'
# Using iloc (position-based)
print(df.iloc[0:3, 0]) # Excludes index 3 (like Python slicing)
loc is label-based (inclusive of the end range).iloc is position-based (exclusive, like Python slicing).