In this article, we’ll take a look at using the Python itertools Module.<\/p>\n\n\n\n
This module is very useful if you want to create different types of iterators suitable for various tasks.<\/p>\n\n\n\n
If you’re able to learn some of the methods of this module, this could be a very useful addition to your toolbox! Let’s get started now, by going through some useful methods.<\/p>\n\n\n\n
In this section, we’ll look at some useful methods which generate iterators.<\/p>\n\n\n\n
To use this module, we must first import it. This is already available in the standard library, so it’s pre-installed!<\/p>\n\n\n
\nimport itertools\n<\/pre><\/div>\n\n\nUsing Python itertools.chain() to chain iterables together<\/h3>\n\n\n\n
The Python itertools.chain()<\/code> method generates an iterator from multiple iterables.<\/p>\n\n\n\nThis simply chains all the iterables together into one sequence and returns a single iterator to that combined sequence.<\/p>\n\n\n\n
The syntax for this method is as follows<\/p>\n\n\n
\niterator = itertools.chain(*sequence)\n<\/pre><\/div>\n\n\nLet’s look at a simple example, to understand this.<\/p>\n\n\n
\nimport itertools\n\nlist1 = ['hello', 'from', 'AskPython']\nlist2 = [10, 20, 30, 40, 50]\ndict1 = {'site': 'AskPython', 'url': 'https:\/\/askpython.com'}\n\n# We can combine lists and dicts (iterables) into a single chain\nfor item in itertools.chain(list1, list2, dict1):\n print(item)\n<\/pre><\/div>\n\n\nHere, we use the iterator directly, by iterating through it, using for item in ...<\/code><\/p>\n\n\n\nOutput<\/strong><\/p>\n\n\n\nhello\nfrom\nAskPython\n10\n20\n30\n40\n50\nsite\nurl\n<\/pre><\/div>\n\n\nHere, although we get the contents of our lists correctly, the dictionary values are not displayed.<\/p>\n\n\n\n
To fix this, we could use dict.items()<\/code> to get a tuple of (key, value)<\/code> pairs.<\/p>\n\n\n\nimport itertools\n\nlist1 = ['hello', 'from', 'AskPython']\nlist2 = [10, 20, 30, 40, 50]\ndict1 = {'site': 'AskPython', 'url': 'https:\/\/askpython.com'}\n\n# We can combine lists and dicts (iterables) into a single chain\nfor item in itertools.chain(list1, list2, dict1.items()):\n print(item)\n<\/pre><\/div>\n\n\nOutput<\/strong><\/p>\n\n\n\nhello\nfrom\nAskPython\n10\n20\n30\n40\n50\n('site', 'AskPython')\n('url', 'https:\/\/askpython.com')\n<\/pre><\/div>\n\n\nIndeed, we now have the values printed as well, using dict1.items()<\/code> as the iterable!<\/p>\n\n\n\nUsing Python itertools.count() to generate a counter-based sequence<\/h3>\n\n\n\n
We can use the function Python itertools.count()<\/code> to make iterators corresponding to a count.<\/p>\n\n\n\niterator = itertools.count(start=0, step=1)\n<\/pre><\/div>\n\n\nHere, this is an iterator which keeps counting indefinitely, from 0 onward.<\/p>\n\n\n\n
This keeps increasing the count by step=1<\/code>. We can also set this to a decimal\/negative number.<\/p>\n\n\n\nFor example, if you want to prove that you have an infinite loop, you can run the below snippet, but it is NOT <\/strong>recommended.<\/p>\n\n\n\nJust make sure that you can understand that itertools.count()<\/code> counts infinitely.<\/p>\n\n\n\nfor num in itertools.count(start=0, step=1):\n # Infinite loop!\n print(num)\n<\/pre><\/div>\n\n\nNow, while you may not find the use of this function immediately obvious, you can combine it with other functions such as the zip method <\/a>to construct sequences.<\/p>\n\n\n\nConsider the below example:<\/p>\n\n\n
\nimport itertools\nnumbers = [100, 200, 300, 400]\n\ndata = list(zip(itertools.count(0, 10), numbers))\n\nprint(data)\n<\/pre><\/div>\n\n\nHere, you can now see the power of iterators! Since iterators produce outputs only on demand, we can zip()<\/code> it with another finite iterable, such as a list!<\/p>\n\n\n\nNow, this is used to construct indices for items in the list, as you can verify using the output!<\/p>\n\n\n
\n[(0, 100), (10, 200), (20, 300), (30, 400)]\n<\/pre><\/div>\n\n\nNow, if you want to have a subset of the iterator sequence using Python itertools.count()<\/code>, you can also use itertools.islice()<\/code> to construct only a slice of the iterator.<\/p>\n\n\n\nimport itertools\n\nfor num in itertools.islice(itertools.count(start=0, step=10), 4):\n print(num)\n\nfor num in itertools.islice(itertools.count(), 0, 50, 10):\n print(num)\n<\/pre><\/div>\n\n\nOutput<\/p>\n\n\n
\n0\n10\n20\n30\n0\n10\n20\n30\n40\n<\/pre><\/div>\n\n\nAs you can observe, both the sequences are identical. This shows that you can have multiple approaches to generate sequences!<\/p>\n\n\n\n
Use the method which you see fit, based on the problem to solve!<\/p>\n\n\n\n
Using itertools.repeat() to repeat a value<\/h2>\n\n\n\n
Suppose you want to repeat a particular value, you can construct an iterator for the repeated value using itertools.repeat(value)<\/code>.<\/p>\n\n\n\nFor example, if you want to construct a sequence of the form (i, 5)<\/code>, where i ranges from 0 to 10, you can use this function!<\/p>\n\n\n\nimport itertools\n\ndata = list(zip(range(10), itertools.repeat(5)))\nprint(data)\n<\/pre><\/div>\n\n\nOutput<\/strong><\/p>\n\n\n\n[(0, 5),\n (1, 5),\n (2, 5),\n (3, 5),\n (4, 5),\n (5, 5),\n (6, 5),\n (7, 5),\n (8, 5),\n (9, 5)]\n<\/pre><\/div>\n\n\nIndeed, we were able to make this sequence easily!<\/p>\n\n\n\n
Another example which this function is useful is if you’re trying to construct squares using map() in Python<\/a>.<\/p>\n\n\n\nsquares = list(map(pow, range(10), itertools.repeat(2)))\nprint(squares)\n<\/pre><\/div>\n\n\nOutput<\/strong><\/p>\n\n\n\n[0, 1, 4, 9, 16, 25, 36, 49, 64, 81]\n<\/pre><\/div>\n\n\nSee how easily we were able to construct it using map()<\/code>?<\/p>\n\n\n\nUsing itertools.tee() to clone sequences<\/h3>\n\n\n\n
There is another useful function called tee()<\/code>, which clones a sequence, and produces two sequences.<\/p>\n\n\n\ncloned1, cloned2 = itertools.tee(original_sequence)\n<\/pre><\/div>\n\n\nThis is based on the Linux tee command, which clones its outputs.<\/p>\n\n\n\n
Here, when you clone a sequence using tee()<\/code>, you cannot use the same iterator again. As a result, you must be very careful when using this function!<\/p>\n\n\n\nimport itertools\n\nsingle_iterator = itertools.islice(itertools.count(), 3)\ncloned1, cloned2 = itertools.tee(single_iterator)\n\nfor num in cloned1:\n print('cloned1: {}'.format(num))\nfor num in cloned2:\n print('cloned2: {}'.format(num))\n<\/pre><\/div>\n\n\nOutput<\/strong><\/p>\n\n\n\ncloned1: 0\ncloned1: 1\ncloned1: 2\ncloned2: 0\ncloned2: 1\ncloned2: 2\n<\/pre><\/div>\n\n\nIndeed, we could see two cloned sequences, having the same outputs!<\/p>\n\n\n\n
Cycle through sequences using itertools.cycle()<\/h3>\n\n\n\n
The itertools.cycle()<\/code> function provides an iterator that we can cycle through indefinitely!<\/p>\n\n\n\nThis is useful if you want to keep switching between states in your application.<\/p>\n\n\n\n
Consider two states of a bulb: “on” and “off”.<\/p>\n\n\n\n
You can construct an iterator which cycles through the two states whenever the switch is pressed!<\/p>\n\n\n
\nimport itertools\n\n# Initially, bulb is switched off, so off is the first element in the list\nbulb_states = itertools.cycle(["off", "on"])\n\nfor _ in range(5):\n # Use next(iterator) to get the current state\n curr_state = next(bulb_states)\n print(f"Bulb state currently {curr_state}")\n<\/pre><\/div>\n\n\nOutput<\/strong><\/p>\n\n\n\nBulb state currently off\nBulb state currently on\nBulb state currently off\nBulb state currently on\nBulb state currently off\n<\/pre><\/div>\n\n\nIndeed, as you can see, the bulb state keeps getting cycled between the two values “on” and “off”!<\/p>\n\n\n\n
Filter items using takewhile() and dropwhile()<\/h3>\n\n\n\n
We can use the Python itertools.takewhile()<\/code> function to filter sequence items as long as a condition is True<\/code>. If the condition becomes False<\/code>, it stops filtering.<\/p>\n\n\n\niterator = itertools.takewhile(condition, *sequence)\n<\/pre><\/div>\n\n\nHere is a simple example, which filters numbers, as long as the number is positive.<\/p>\n\n\n
\nimport itertools\n\nsequence = itertools.takewhile(lambda x: x > 0, [1, 2, 3, -1, 10])\n\nfor item in sequence:\n print(item)\n<\/pre><\/div>\n\n\nOutput<\/strong><\/p>\n\n\n\n1\n2\n3\n<\/pre><\/div>\n\n\nHere, the sequence stopped after 3, since the next element is -1.<\/p>\n\n\n\n
Similarly, the itertools.dropwhile()<\/code> filters elements as long as a condition is False<\/code> and returns all the elements after the first non-false value.<\/p>\n\n\n\nimport itertools\n\ndata = itertools.dropwhile(lambda x: x < 5, [3, 12, 7, 1, -5])\nfor item in data:\n print(item)\n<\/pre><\/div>\n\n\nOutput<\/strong><\/p>\n\n\n\n12\n7\n1\n-5\n<\/pre><\/div>\n\n\nConstruct combinations using combinations()<\/h3>\n\n\n\n
We can also construct combination sequences using Python itertools.combinations()<\/code>.<\/p>\n\n\n\niterator = itertools.combinations(*sequence, r)\n<\/pre><\/div>\n\n\nHere is a simple example:<\/strong><\/p>\n\n\n\nimport itertools\nwords = ['hello', 'from', 'AskPython', 'how']\nresults = itertools.combinations(words, 2)\nfor item in results:\n print(item)\n<\/pre><\/div>\n\n\nOutput<\/strong><\/p>\n\n\n\n('hello', 'from')\n('hello', 'AskPython')\n('hello', 'how')\n('from', 'AskPython')\n('from', 'how')\n('AskPython', 'how')\n<\/pre><\/div>\n\n\nIf you want to have repetitions of consecutive elements in the combinations, you can use combinations_with_replacement()<\/code>.<\/p>\n\n\n\nresults = itertools.combinations_with_replacement(words, 3)\n\nfor item in results:\n print(item)\n<\/pre><\/div>\n\n\nOutput<\/strong><\/p>\n\n\n\n('hello', 'hello', 'hello')\n('hello', 'hello', 'from')\n('hello', 'hello', 'AskPython')\n('hello', 'hello', 'how')\n('hello', 'from', 'from')\n('hello', 'from', 'AskPython')\n('hello', 'from', 'how')\n('hello', 'AskPython', 'AskPython')\n('hello', 'AskPython', 'how')\n('hello', 'how', 'how')\n('from', 'from', 'from')\n('from', 'from', 'AskPython')\n('from', 'from', 'how')\n('from', 'AskPython', 'AskPython')\n('from', 'AskPython', 'how')\n('from', 'how', 'how')\n('AskPython', 'AskPython', 'AskPython')\n('AskPython', 'AskPython', 'how')\n('AskPython', 'how', 'how')\n('how', 'how', 'how')\n<\/pre><\/div>\n\n\nSimilarly, you could list the permutations using permutations()<\/code> and permutations_with_replacement()<\/code>.<\/p>\n\n\n\nThat concludes some of the important functions of this module. For more functions, you can consult the official documentation<\/a>.<\/p>\n\n\n\n
\n\n\n\nConclusion<\/h2>\n\n\n\n
In this article, we looked at various functions in the Python itertools<\/code> module. Depending on your problem, you could use one of multiple approaches, to construct sequences quickly!<\/p>\n\n\n\nReferences<\/h2>\n\n\n\n