Memory-mapped arrays for large data in NumPy - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is a memory-mapped array in NumPy?

A memory-mapped array is a way to access large data stored on disk as if it were in memory, without loading the entire data into RAM. It allows working with data too big to fit in memory by loading only needed parts.

beginner

How do you create a memory-mapped array in NumPy?

Use numpy.memmap(filename, dtype, mode, shape) to create a memory-mapped array. It links the array to a file on disk, specifying data type, access mode, and shape.

beginner

Why use memory-mapped arrays instead of loading the whole file into memory?

Memory-mapped arrays let you work with files larger than your RAM by loading only parts you need. This saves memory and speeds up processing for big data.

intermediate

What does the mode parameter in numpy.memmap control?

The mode controls how the file is opened: 'r' for read-only, 'r+' for read-write, 'w+' to create or overwrite, and 'c' for copy-on-write.

intermediate

Can you modify data in a memory-mapped array?

Yes, if you open the memory-mapped array in a write mode like 'r+' or 'w+', you can change data and it will be saved back to the file on disk.

What is the main advantage of using a memory-mapped array?

A It speeds up CPU calculations by using GPU

B It allows working with data larger than RAM by loading parts on demand

C It automatically cleans data errors

D It compresses data to save disk space

Memory-mapped arrays let you access large files without loading all data into RAM, improving memory use.

Which NumPy function creates a memory-mapped array?

A numpy.memmap()

B numpy.loadtxt()

C numpy.array()

D numpy.fromfile()

numpy.memmap() creates a memory-mapped array linked to a file.

If you want to modify a memory-mapped array and save changes, which mode should you use?

A 'r+'

B 'c'

C 'r'

D 'r--'

Mode 'r+' opens the file for reading and writing, allowing changes to be saved.

What happens if you open a memory-mapped array with mode 'c'?

A Overwrite the file

B Read-only access

C Copy-on-write, changes not saved to disk

D Append data to the file

Mode 'c' creates a copy-on-write mapping; changes are not saved back to the file.

Which of these is NOT a benefit of memory-mapped arrays?

A Handling files larger than RAM

B Saving memory by loading data on demand

C Faster access to small parts of large files

D Automatic data cleaning

Memory-mapped arrays do not clean data; they only manage how data is loaded and accessed.

Explain how memory-mapped arrays help when working with very large datasets.

Think about how you might read a huge book by opening only the pages you need.

Describe the steps to create and use a memory-mapped array in NumPy.

Remember the function and what parameters it needs.