Python-like generator for reading large CSV file?

I have a large csv file, typically I just read everything into a array like this

num =csvread(filename,2,2)

if csv file is big, the num array will be big as well and sometime it blowout my memory.

In python, there is a way to read the file but without holding the data in the memory. the methods use generator object. For example,

Base case(w/o using generator object):

value=[len(x) for x in open('filename.tx')] this case will hold the data in the memory

another case(use generator object:

it=(len(x) for x in open('filename.tx'))

the it is a generator object and it only hold the memory size for the generator expression without holding memory for all the data, later, we just extract the data by using next(it).

I am wondering if Matlab has similar generator object as Python when reading csv file to save the memory? Thanks

Answer 1

[Guillaume]

Use datastores and tall arrays.

Answer 2

[roudan]

Thank you, Guillaume. That is awesome! I appreciate it.