Float types (float16, float32, float64) in NumPy - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is a float16 type in numpy?

Float16 is a 16-bit floating-point number type in numpy. It uses less memory but has lower precision and smaller range compared to float32 and float64.

beginner

How does float32 differ from float64 in numpy?

Float32 uses 32 bits to store a floating-point number, offering a balance between memory use and precision. Float64 uses 64 bits, providing higher precision and a larger range but uses more memory.

intermediate

Why choose float16 over float64 in numpy?

Choose float16 when memory is limited and you can accept lower precision. It is useful for large arrays where saving memory is more important than exact precision.

intermediate

What happens if you perform calculations mixing float16 and float64 in numpy?

Numpy will upcast to the higher precision type (float64) to avoid losing precision during calculations.

beginner

How can you check the data type of a numpy array?

Use the `.dtype` attribute of the numpy array to see its data type, for example: `array.dtype`.

Which numpy float type uses the least memory?

A float16

B float32

C float64

D float128

Float16 uses 16 bits, which is less memory than float32 (32 bits) and float64 (64 bits).

What is the main advantage of float64 over float32?

A Uses less memory

B Higher precision and larger range

C Smaller range

D Faster calculations

Float64 provides higher precision and can represent larger or smaller numbers than float32.

If you add a float16 and a float64 numpy array, what will be the result type?

A float16

B float32

C float64

D int

Numpy upcasts to the higher precision type, so the result will be float64.

Which float type is best for saving memory when precision is less important?

A float64

B int32

C float32

D float16

Float16 uses the least memory and is suitable when precision can be sacrificed.

How do you find the data type of a numpy array named 'arr'?

A arr.dtype

B arr.type()

C type(arr)

D arr.data_type

The `.dtype` attribute shows the data type of a numpy array.

Explain the differences between float16, float32, and float64 in numpy.

Think about how memory and precision trade off among these types.

Describe what happens when numpy performs operations on arrays with different float types.

Consider how numpy avoids losing precision in mixed-type calculations.