Pandas Series & DataFrame Internals (.loc vs .iloc)

1Concept

Pandas structures data into 1D `Series` and 2D `DataFrame` objects. `.loc` accesses rows and columns by label/name; `.iloc` accesses elements strictly by integer 0-indexed positions. Understanding the distinction prevents subtle data extraction bugs.

2Architecture Diagram

df.loc['row_label', 'col_name']  ---> Label-based indexing (inclusive of end!)
df.iloc[0:2, 0:3]                 ---> Position-based integer indexing (exclusive of end!)

3Code Example

# Simulating DataFrame index behaviors
records = [
    {"user_id": 101, "name": "Alice", "spend": 450.0},
    {"user_id": 102, "name": "Bob",   "spend": 120.5}
]

print("=== DataFrame Indexing Hierarchy ===")
print(f"Record 0 (iloc[0]): {records[0]['name']} (Spend: ${records[0]['spend']})")
print("Rule: Use .loc for explicit labels; use .iloc for integer row indices.")

4Expected Output

=== DataFrame Indexing Hierarchy ===
Record 0 (iloc[0]): Alice (Spend: $450.0)
Rule: Use .loc for explicit labels; use .iloc for integer row indices.

5Key Takeaways

• ✓`.loc` slices are INCLUSIVE of both start and end bounds; `.iloc` slices are EXCLUSIVE of end.
• ✓Avoid chained indexing (`df['col'][0] = val`); it triggers `SettingWithCopyWarning`.
• ✓Use `.copy()` when slicing a DataFrame to avoid mutating the original dataset.