iloc for position-based selection in Pandas - Step-by-Step Execution

Concept Flow - iloc for position-based selection

Start with DataFrame

↓

Use iloc with positions

↓

Select rows/columns by integer index

↓

Return subset DataFrame or Series

↓

Use or display selected data

iloc lets you pick rows and columns by their integer position in the DataFrame, not by labels.

Execution Sample

Pandas

import pandas as pd

df = pd.DataFrame({
  'A': [10, 20, 30],
  'B': [40, 50, 60]
})

subset = df.iloc[1:3, 0:2]

This code selects rows 1 and 2, and columns 0 and 1 by position from the DataFrame.

Execution Table

Step	Action	Row positions selected	Column positions selected	Resulting DataFrame
1	Create DataFrame	-	-	[[10, 40], [20, 50], [30, 60]]
2	Apply iloc[1:3, 0:2]	Rows 1 and 2	Columns 0 and 1	[[20, 50], [30, 60]]
3	Display subset	-	-	DataFrame with rows index 1 and 2, columns 'A' and 'B'

💡 Selection complete, subset DataFrame returned with specified rows and columns by position.

Variable Tracker

Variable	Start	After iloc selection
df	[[10, 40], [20, 50], [30, 60]]	[[10, 40], [20, 50], [30, 60]]
subset	Not defined	[[20, 50], [30, 60]]

Key Moments - 3 Insights

Why does iloc use numbers instead of column names?

iloc always uses integer positions to select rows and columns, not labels. This is shown in the execution_table step 2 where positions 1:3 and 0:2 are used.

What happens if I use iloc with a position outside the DataFrame size?

It will raise an IndexError because iloc expects valid integer positions within the DataFrame size. This is not shown here but is important to remember.

Does iloc include the stop index in slicing?

No, iloc slicing excludes the stop index, like normal Python slicing. Step 2 shows rows 1 and 2 selected, but not row 3.

Visual Quiz - 3 Questions

Look at the execution_table, what rows does iloc select at step 2?

A Rows 0 and 1

B Rows 2 and 3

C Rows 1 and 2

D Rows 0, 1, and 2

💡 Hint

Check the 'Row positions selected' column in execution_table step 2.

According to variable_tracker, what is the value of 'subset' after iloc selection?

A [[10, 40], [20, 50], [30, 60]]

B [[20, 50], [30, 60]]

C [[40, 50], [60, 70]]

D Not defined

💡 Hint

Look at the 'After iloc selection' value for 'subset' in variable_tracker.

If we change iloc to df.iloc[0:2, 1:2], what columns will be selected?

A Column 1 only

B Column 0 only

C Columns 0 and 1

D Columns 1 and 2

💡 Hint

Recall that iloc slicing excludes the stop index; see concept_flow and execution_table for slicing behavior.

Concept Snapshot

iloc selects rows and columns by integer position.
Syntax: df.iloc[row_start:row_stop, col_start:col_stop]
Slices exclude the stop index.
Returns a DataFrame or Series subset.
Positions start at 0, not labels.

Full Transcript

We start with a DataFrame of 3 rows and 2 columns labeled 'A' and 'B'. Using iloc, we select rows by their integer positions 1 and 2, and columns 0 and 1. iloc uses integer positions, not labels, so the first row is position 0. The slicing excludes the stop index, so 1:3 means rows 1 and 2 only. The resulting subset DataFrame contains the selected rows and columns. Variables 'df' and 'subset' track the original and selected data. Common confusions include iloc using positions not labels, slicing excluding the stop index, and errors if positions are out of range. The visual quiz checks understanding of row selection, variable values, and column slicing behavior.