Function definitions vs Lambda expression

Question

Given these two snippets:

def sex (code):
    return {'m': 'masculino', 'f': 'femenino', '': 'ignorado'} [code]

and

sex = lambda code: {'m': 'masculino', 'f': 'femenino', '': 'ignorado'} [code]

What is the actual difference? Do they have different behaviour?

And above all: Is one preferrable over the other?

I ask this, because one of my answers in this forum ( python3 - learning about searching, this very simple example does not work right ), which offered both possibilities, got truncated by another user ("revised"), deleting the lambda term. The reason for the revision was that the last option (the lambda expression) is worse than the function definition. This made me wonder:

Why are lambda expressions worse than function definitions? Especially simple ones like these.

BrenBarn · Accepted Answer · 2012-12-13 05:21:20Z

7

There is no functional difference between a lambda and a def that returns the same thing. They are exactly the same and have identical bytecode.

>>> def sex (code):
...     return {'m': 'masculino', 'f': 'femenino', '': 'ignorado'} [code]
>>> sex2 = lambda code: {'m': 'masculino', 'f': 'femenino', '': 'ignorado'} [code]
>>> sex.__code__.co_code == sex2.__code__.co_code
True

I would say it is wrong to say that either way of writing functions is "worse" in a general sense. In your example, however, the def version is better. The reason is that in general the purpose of lambdas is to concisely write short, anonymous functions within expressions. There is no real point to doing name = lambda: .... There's no reason to define an anonymous function if you're going to immediately give it a name.

The point of lambdas is that you can do someFuncRequiringCallback(lambda x: x+1), where you don't want to give the function a name or use it for anything besides this situation. If you do want to use a function across multiple contexts, just define it with a regular def.

edited Dec 13, 2012 at 5:21

answered Dec 13, 2012 at 5:15

BrenBarn

253k39 gold badges421 silver badges392 bronze badges

Sign up to request clarification or add additional context in comments.

3 Comments

DSM Over a year ago

I can think of one minor difference: print sex.__name__, sex2.__name__.

BrenBarn Over a year ago

@DSM: That's no more a difference than doing def foo: return 1 and def bar: return 1. The two functions are still the same.

DSM Over a year ago

Since the functions have the same code I completely agree that there's no functional difference, but there have been a few times when I've checked __name__, and I can't do that with a lambda. YMMV.

Abhijit · Accepted Answer · 2012-12-13 05:54:54Z

They are the same, check the disassembled python byte code

>>> def sex1 (code):
    return {'m': 'masculino', 'f': 'femenino', '': 'ignorado'} [code]

>>> sex2 = lambda code: {'m': 'masculino', 'f': 'femenino', '': 'ignorado'} [code]
>>> dis.dis(sex1)
  2           0 BUILD_MAP                3
              3 LOAD_CONST               1 ('masculino')
              6 LOAD_CONST               2 ('m')
              9 STORE_MAP           
             10 LOAD_CONST               3 ('femenino')
             13 LOAD_CONST               4 ('f')
             16 STORE_MAP           
             17 LOAD_CONST               5 ('ignorado')
             20 LOAD_CONST               6 ('')
             23 STORE_MAP           
             24 LOAD_FAST                0 (code)
             27 BINARY_SUBSCR       
             28 RETURN_VALUE        
>>> dis.dis(sex2)
  1           0 BUILD_MAP                3
              3 LOAD_CONST               1 ('masculino')
              6 LOAD_CONST               2 ('m')
              9 STORE_MAP           
             10 LOAD_CONST               3 ('femenino')
             13 LOAD_CONST               4 ('f')
             16 STORE_MAP           
             17 LOAD_CONST               5 ('ignorado')
             20 LOAD_CONST               6 ('')
             23 STORE_MAP           
             24 LOAD_FAST                0 (code)
             27 BINARY_SUBSCR       
             28 RETURN_VALUE

Even both of them have the same type

>>> type(sex1)
<type 'function'>
>>> type(sex2)
<type 'function'>

Community · Accepted Answer · 2017-05-23 12:29:55Z

3

In addition to the information provided in BrenBarn's excellent answer, you will find many Python programmers who insist on always using def over lambda, so I will attempt to explain why that is.

It basically comes down to the following line from the Zen of Python:

There should be one-- and preferably only one --obvious way to do it.

This is referenced in Guido van Rossum's post about this issue (Guido is the creator of Python):

Why drop lambda? Most Python users are unfamiliar with Lisp or Scheme, so the name is confusing; also, there is a widespread misunderstanding that lambda can do things that a nested function can't -- I still recall Laura Creighton's Aha!-erlebnis after I showed her there was no difference! Even with a better name, I think having the two choices side-by-side just requires programmers to think about making a choice that's irrelevant for their program; not having the choice streamlines the thought process. Also, once map(), filter() and reduce() are gone, there aren't a whole lot of places where you really need to write very short local functions; Tkinter callbacks come to mind, but I find that more often than not the callbacks should be methods of some state-carrying object anyway (the exception being toy programs).

edited May 23, 2017 at 12:29

CommunityBot

11 silver badge

answered Dec 13, 2012 at 5:31

Andrew Clark

210k36 gold badges285 silver badges310 bronze badges

1 Comment

Hyperboreus Over a year ago

Thank you for your answer and the link to Guido's post. Very interesting. And I experience it myself: Since we moved from py2.7 to py3.2, most map(), filter(), etc expressions became generators like (f(x) for x in l if x). Nevertheless I am happy we still got lambda.

Ry- · Accepted Answer · 2012-12-13 05:15:59Z

0

Simple lambdas aren't necessarily worse than nested functions (and no, there's no difference between those two examples but syntax), but this particular case can be even simpler:

sex = {'m': 'masculino', 'f': 'femenino', '': 'ignorado'}

Then you use sex.__getitem__.

answered Dec 13, 2012 at 5:15

community wiki

Ry-

Comments

aquavitae · Accepted Answer · 2012-12-13 05:33:34Z

0

They have the same behaviour, but the key difference is that lambda is an expression and as such can't contain statements. lambdas are not necessarily worse that function definitions, provided they are used appropriately. However, they are often are used to write one-line functions simply because they can, even if it makes the code practically impossible tor read. For example, consider the following:

sortfilter = lambda a, b, c: sorted(filter(lambda i: i % 2, (i for i in a if i in b)), key=lambda v: c.get(v, v)))

Possible to write in a one-line lambda, but a really bad idea!

answered Dec 13, 2012 at 5:33

aquavitae

19.4k12 gold badges68 silver badges110 bronze badges

Comments

user3412656 · Accepted Answer · 2015-05-21 10:32:07Z

0

It helps differentiate between things. If I am defining a function I will start with def, however inside of that function I will usually use lambda.

It is the same concept as writing in english. If a quote is quoting a quote you will want to change the looks of the nested quote. This is also true for parenthesis.

answered May 21, 2015 at 10:32

user3412656

Collectives™ on Stack Overflow

Function definitions vs Lambda expression

6 Answers 6

3 Comments

Comments

1 Comment

Comments

Comments

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

6 Answers 6

3 Comments

Comments

1 Comment

Comments

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related