[Python]Generators

Definition

Generators are special python objects that have the following features:

iteratable. so one can use next() or with for loop to get the elements.
return elements when asked, one element each time; unlike list, the elements are not generated until asked for.
the local state is saved after each call/request, so it can resume when next call arrives.

Advantages

Because of the above features, generators are more memory efficient than lists, and thus used in handling large datastream.

How to construct

There are two ways to make generators, summarized in the following table:

Category	Explanation
Generator functions	use yield other than return to return a value from a function.
Generator expression	replace the surrounding ‘[]’ of list comprehension with ‘()’.

Below are examples for each of them:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16


# Example 1: generator functions
def countdown(num):
    print('Starting')
    while num > 0:
        yield num
        num -= 1
a=countdown(10) # returned is a generator
next(a) # get next value
for i in a:
  print(a) # print each element

# Example 2: generator expression
mylist = ['a', 'b', 'c', 'd']
g = (x+’ZZG’ for x in mylist) # a generator
for s in g:
  print(s) # print each element

In the above first example, yield is used to return a value, and in the second example, a generator is constructed in the format of ‘(xx for xx in an-iterator)’.

Notes

A generator object can only be looped over once. To loop again, regenerate it.
Generators are more efficient than list only when the dataset size is near or over the limit of momery.

Enjoy programming 😃

References

https://realpython.com/introduction-to-python-generators/

Last modified on 2018-10-17