Main Content

Multiple identical keys in a Python dict - yes, you can!

Archive - Originally posted on "The Horse's Mouth" - 2012-11-24 07:20:30 - Graham Ellis

If you have a list, you can only have one element at each number - there's just a single positon [0], a single [1] and so on. That's clear, and natural to understand. With a dict in Python (like a hash in Perl or an associative array in PHP), it's peceived wisdom that you can only have one element with a paricular name.

Let's see an example - setting up a dict:
  control = {"Andrew" : "Cambridge", "Barabara" : "Bloomsbury", "Andrew": "Corsica"}
  print control

Which sets up a dict called "control" ... adds an element called Andrew, another element called Barbara, and then replaces the original Andrew element with a new element of the same name. So that although we've added three elements, we only have two when we print it out:
  {'Barabara': 'Bloomsbury', 'Andrew': 'Corsica'}

But ... what if I wanted to store multiple "Andrew"s? In Python, if I use mutable objects as my keys, then I can create multiple identical objects, and each can be a separate key. Let's use a very simple object - with just a single member that's a string:
  class person(object):
    def __init__(self,name):
      self.name = name

I can then set up my dict as follows:
  alternate = {person("Andrew") : "Cambridge", person("Barabara") : "Bloomsbury", person("Andrew"): "Corsica"}
  print alternate
and when I print out the results, I do get two (identical but different) Andrews:
  {'Barabara': 'Bloomsbury', 'Andrew': 'Cambridge', 'Andrew': 'Corsica'}

Newcomers to this sort of structure worry about how they can access individual elements once the dict has been set up. That's understandable but not really a problem - if two key objects are identical, it doesn't matter which one you refer to. And if they're not identical, then you can differentiate between the keys somehow. And you can loop through all the keys via the keys method:
  for staff in alternate.keys():
    print staff, "lives in", alternate[staff]

If you refer to an element with an existing name, though, you now need to be careful. For example:
  alternate[person("Andrew")] = "Tignabruiach"
Will add yet another "Andrew" to our dict.

There are further examples within the source code from which the above examples are taken [here].



A further approach to the need to multiple identical keys is to store a list at each member of a dict. There's a new example of setting one of these up [here].

In my example, I've provided an add method to store new data, and I would need to use a loop to access members with the same key.

Examples from our new Intermediate Python course which ran for the first time last week.