Python things you never need: Empty lambda functions

Ever wanted to create an empty Python function in-line?
No?
Me neither.
But my coworker does...
We found the answer:
<pre>
x = lambda : None
</pre>

If you ever needed this, please tell me ;)

[Edit]
Apparently some people do need it :)
Btw, most people seem to try
Most people try
lambda: pass
first.
But you have to remember, the thing at the right side of the colon is the return value of the function, not the body! So an expression is needed, not a statement.
[\Edit]

Comments

  1. UnknownJuly 13, 2011 at 9:44 AM

    Thanks! Yes. I needed it. If you have a better suggestion, I'm all ears, but my scenario involves me trying to discover Autodesk Maya's next script job number and to do so I can't rely on the most recent script job. Instead, I have to create a new script job, get its ID and then just delete it. Thing is, I don't actually want this script job to do anything, but it *requires* a function. So, I need an empty lambda. lambda:pass was improper syntax, but lambda:None works beautifully. Thanks!

    ReplyDelete
  2. Andreas MuellerJuly 13, 2011 at 10:22 AM

    Yeah, I know that there are some situations where you might need it. It just seems kind of weired ;)

    ReplyDelete
  3. AnonymousMarch 7, 2012 at 1:12 AM

    Thanks! I needed this as well :)

    ReplyDelete
  4. John EApril 6, 2012 at 8:02 PM

    I wanted (using the verb 'need' implies there is no other way around this, which I am unsure of) to do this in order to give an empty function to threading.Timer. Of course lambda:0 or any expression that does not mess with program or system memory (maybe other considerations?) would have done the trick, but sometimes it is nice to use the official null object identifier in order to clarify intent.

    ReplyDelete
  5. John EApril 6, 2012 at 8:05 PM

    Technically, it shouldn't ever truly be "necessary" but it is more desirable when the cost of changing source or somebody else's (or your own) code would take way too long. In my case, I wasn't about to modify the threading module. Although there may have been some other workaround.

    ReplyDelete
  6. Munciun CaractarJune 9, 2012 at 7:19 AM

    I just used this. I have a dictionary full of lambda functions, and so I did

    some_dict.get(the_key, lambda: None)(), which seems to work beautifully.

    ReplyDelete
  7. Mr. ThyOctober 3, 2012 at 11:32 PM

    One of my methods has an optional argument that is a callback function. By default it should do nothing, so I needed to test if it was None or not in the function, and then call the callback or not. By defining my argument as "callback = lambda x : None" I eliminate this bothersome test.

    ReplyDelete
  8. AnonymousFebruary 27, 2013 at 9:27 PM

    As far as I know, this is what we call a special case object. So, I disagree with John E on that 'it shouldn't ever truly be "necessary"'. It is as necessary as empty lists, empty sets, empty iterators, empty queues, emptry trees, zeros, ones (neutral element of multiplication) etc.
    The emtpy set for instance is the very foundation of every single piece of math we know of today (cf. http://www.science.uva.nl/~seop/entries/set-theory/primer.html). So, declaring this SCO a something unnecessary is quite naive.

    Thanks anyway for this pyhton idiom. :) However, the title could be more constructive. ;)

    ReplyDelete
    Replies
    1. Andreas MuellerApril 2, 2013 at 2:38 PM

      I mostly agree with you. The title was never meant to be very serious ;)

      The reason that the empty function feels more artificial in Python than an empty list is maybe because I am not a functional programmer (and it is debatable in how far it is possible to write functional code in Python).

      So functions don't really feel to me like "just another object" - not claiming that this isn't naive ;)

      Delete
    2. AnonymousMay 14, 2013 at 8:35 PM

      I needed this today when writing testing code; I needed to replace a function in a class I was testing with a do-nothing function.

      Delete
    3. VekyApril 18, 2015 at 8:40 PM

      "empty function" is _not_ the same as the rest of your examples. Your "special case object" is in fact neutral element: empty list for concatenation, empty set for union, empty iterator for chaining, zeros for addition, ones for multiplication.

      In that sense, "neutral function" is in fact identity: lambda x:x, neutral for composition. lambda:None is something else, kinda like _terminal_ object (something like NaN for numbers) but not quite.

      Delete
  9. AnonymousJune 2, 2013 at 9:18 PM

    There are a myriad of circumstances requiring a null function!

    def nul(): pass;

    Is perfectly correct python. Following x = nul() (x is None) is True.
    Curiously lambda : pass produces a syntax error! Therefore your solution is pythonicly correct.

    As a matter of information python is a first rate language for functional programming, particularly higher order functions.

    A higher order function is one that accepts a function or functions as parameter and/or returns a function as result.

    See here http://docs.python.org/2/library/functools.html for some basic higher order functional techniques.

    ReplyDelete
    Replies
    1. Andreas MuellerJune 2, 2013 at 9:37 PM

      The reason that "pass" throws a syntax error is actually quite simple: lambda requires an expression, not a statement. "pass" is a statement.

      Delete
    2. Andreas MuellerJune 2, 2013 at 9:38 PM

      Which I already said in the post btw ^^

      Delete
  10. AnonymousJune 7, 2014 at 8:32 AM

    I needed this, thanks :)

    ReplyDelete
  11. UnknownAugust 18, 2014 at 11:53 PM

    This comment has been removed by the author.

    ReplyDelete
  12. AnonymousOctober 6, 2014 at 2:31 PM

    Quite useful for me as a default implementation of some expected interface.

    ReplyDelete
  13. AnonymousOctober 29, 2014 at 11:33 AM

    I've used it as a mock object for error printing during testing:
    def run_test(function, params, expected, logger = print):
    ...

    try:
    run_test(lambda x : x * 2, 8, 32, lambda x : None)
    except RuntimeError:
    pass

    ReplyDelete

Post a Comment

Popular posts from this blog

Machine Learning Cheat Sheet (for scikit-learn)

Image
As you hopefully have heard, we at scikit-learn are doing a user survey (which is still open by the way). One of the requests there was to provide some sort of flow chart on how to do machine learning. As this is clearly impossible, I went to work straight away. This is the result: [edit2] clarification: With ensemble classifiers and ensemble regressors I mean random forests , extremely randomized trees, gradient boosted trees , and the soon-to-be-come weight boosted trees (adaboost). [/edit2] Needless to say, this sheet is completely authoritative.
Read more

A Wordcloud in Python

Image
Last week I was at Pycon DE , the German Python conference. After hacking on scikit-learn a lot last week, I decided to to something different on my way back, that I had planned for quite a while: doing a wordl -like word cloud . I know, word clouds are a bit out of style but I kind of like them any way. My motivation to think about word clouds was that I thought these could be combined with topic-models to give somewhat more interesting visualizations. So I looked around to find a nice open-source implementation of word-clouds ... only to find none. (This has been a while, maybe it has changed since). While I was bored in the train last week, I came up with this code . A little today-themed taste:
Read more

MNIST for ever....

Image
[update] This post is a bit old, but many people still seem interested. So just a short update: Nowadays I would use Python and scikit-learn to do this. Here is an example of how to do cross-validation for SVMs in scikit-learn. Scikit-learn even downloads MNIST for you. [/update] MNIST is, for better or worse, one of the standard benchmarks for machine learning and is also widely used in then neural networks community as a toy vision problem. Just for the unlikely case that anyone is not familiar with it: It is a dataset of handwritten digits, 0-9, in black on white background. It looks something like this: There are 60000 training and 10000 test images, each 28x28 gray scale. There are roughly the same number of examples of each category in the test and training datasets. I used it in some papers myself even though there are some reasons why it is a little weird. Some not-so-obvious (or maybe they are) facts are: - The images actually contain a 20x20 patch of digi
Read more