27rohitb.github.io

Advanced Concepts

This file has some advanced concepts like Generators, Iterators etc.

Generators:

This is memory efficient way to implement a function or a forloop in a function.
The keywords used are : yield and next:

NOTE:

• Generators can return ONLY 1 item, so they should be used in functions “generating” some kind of data, and NOT in function that are manipulating it.

Taking example of function that computes the square of numbers in a given list.

# Normal, less mem-efficient function:

def sq_num(alist):
   my_result = []
   for i in alist:
      my_result.append(i * i)
   return my_result
   
# Usage:    print(sq_num([1,2,3,4]))

# Using Generators (mem-efficient approach):

def sq_num(alist):
   for i in alist:
      yield (i * i)  # yield KEYWORD used HERE
      
            res = sq_num([1,2,3,4])
# Usage:    
print(res)        # will return a generator object
print(next(res))  # NEXT keyword used HERE, will give output for 1st value in list, i.e "1"

Converting list comprenhension to generator:

list_comp = [x*x for x in my_list]
my_gen = (x*x for x in my_list)

Converting Generator to list :

my_list = list(my_gen)

NOTE: Almost ALL dunder(__) methods have a normal method which calls them internally.

if ‘name’ == ‘main’ :

The above important. When we import a module, everything not under the above line run by the python. The function under the above file wont be executed!!.

Also note that, when we run a module directly, the __name__ is directly set as __main__ by default, but when it is being imported, it is set to <file name of the module> by default.

Iterators & Iterables:

Iterable- Something that can be looped over. For anything to be iterable, it needs to have __iter__ . So, Iterable returns an “iterator”;

An iterator is an object with a state, which remember where it is, during the iteration. They also know, how to get their next value, that is using __next__ method. Note: Iterators are also iterables (can be looped over).

NOTE :

• Iterators can ONLY go forward or reset it!
• Iterators don’t need to end, as long they have next value.

Example of an iterator class :

class MyRange:

   def __init__(self, start, end):
      self.value = start
      self.end = end
   
   def __iter__(self):
      return self
   
   def __next__(self):
      #create some sort of recursive loop here
      #
      # Break condition;
      if self.value >= self.end:
         raise StopIteration
         
      # save the state, that be returned later
      current = self.value
      
      # update the state, internally
      self.value += 1
      
      # finally return the saved state.
      return current

Difference between Generators & iterators:

Generators are also iterators, but the __iter__ & __next__ methods are created automatically.

Also, Geneator -> for a function, while Iterator -> for a class.