Type promotion in operations in NumPy - Step-by-Step Execution

Concept Flow - Type promotion in operations

Start with arrays of different types

↓

Perform operation (e.g., addition)

↓

Check types of operands

↓

Apply type promotion rules

↓

Result array has promoted type

↓

Output result with promoted type

When numpy arrays of different types are used in operations, numpy promotes their types to a common type that can hold all values safely before computing the result.

Execution Sample

NumPy

import numpy as np
arr1 = np.array([1, 2, 3], dtype=np.int32)
arr2 = np.array([1.5, 2.5, 3.5], dtype=np.float64)
result = arr1 + arr2
print(result)
print(result.dtype)

Adds an int32 array and a float64 array, showing how numpy promotes types to float64 for the result.

Execution Table

Step	Action	Operand 1 Type	Operand 2 Type	Promoted Type	Result Value	Result Type
1	Create arr1	int32	-	-	[1 2 3]	int32
2	Create arr2	-	float64	-	[1.5 2.5 3.5]	float64
3	Add arr1 + arr2	int32	float64	float64	[2.5 4.5 6.5]	float64
4	Print result	-	-	-	[2.5 4.5 6.5]	float64
5	Print result dtype	-	-	-	float64	dtype('float64')

💡 Operation completes with result promoted to float64 because float64 can hold int32 and float values safely.

Variable Tracker

Variable	Start	After Operation	Final
arr1	[1 2 3] (int32)	[1 2 3] (int32)	[1 2 3] (int32)
arr2	[1.5 2.5 3.5] (float64)	[1.5 2.5 3.5] (float64)	[1.5 2.5 3.5] (float64)
result	N/A	[2.5 4.5 6.5] (float64)	[2.5 4.5 6.5] (float64)

Key Moments - 2 Insights

Why is the result array's type float64 instead of int32?

Because numpy promotes types to the more general type that can hold all values safely. Here, float64 can hold both int32 and float64 values, so the result is float64 (see execution_table step 3).

Does the original int32 array change its type after the operation?

No, the original arrays keep their types unchanged. Only the result array has the promoted type (see variable_tracker for arr1 and result).

Visual Quiz - 3 Questions

Look at the execution table at step 3, what is the promoted type for the addition?

A float64

B int32

C int64

D object

💡 Hint

Check the 'Promoted Type' column at step 3 in the execution_table.

At which step does the result array get created with the promoted type?

A Step 2

B Step 3

C Step 1

D Step 4

💡 Hint

Look for the step where the addition happens and the result type is assigned.

If arr2 was int16 instead of float64, what would be the promoted type when adding to arr1 (int32)?

A int16

B float64

C int32

D object

💡 Hint

Consider numpy's type promotion rules for integer types of different sizes.

Concept Snapshot

Type promotion in numpy operations:
- When arrays of different types are combined, numpy promotes to a common type.
- The promoted type can hold all values without loss.
- Example: int32 + float64 -> float64 result.
- Original arrays keep their types unchanged.
- Result array has the promoted type.

Full Transcript

This visual execution shows how numpy promotes types when performing operations on arrays of different types. Starting with an int32 array and a float64 array, numpy adds them by promoting the int32 to float64 to avoid losing decimal information. The result array is float64 type. The original arrays remain unchanged. This ensures safe and accurate calculations.