Question

I don’t think this is a duplicate, but I just might not be using the correct terminology in my searching. My apologies if this is the case.

What I have is a dictionary mapping a tuple (custom object, string) to a float. To be exact then, it is a 2-tuple. What I am trying to do is get back a representation of all entries in the dictionary that match on the custom object.

Example:

mydict[(co1, 'hello')] = 0.01
mydict[(co2, 'bye')] = 0.02

Where co1 and co2 are supposed to represent two distinct custom objects. I want to find all entries that contain co1 (it’s logical equivalent) in the tuple which is the key.

So my question then is how to reduce this 2_tuple=>float dictionary down to a string=>float dictionary when the string is the unique (non-matched) tuple.

What I have tried:

for custom in custom_object:
    for k, v in mydict.iteritems():
        if custom in k:
           #store this particular entry into another data structure or otherwise process

Answer 1

You can use unpacking in a comprehension:

dict((s, v) for (o, s), v in mydict.iteritems() if o is co1)

From Python 2.7:

{s: v for (o, s), v in mydict.iteritems() if o is co1}