Are there a thing call "semi-monad" or "counter-monad"? - haskell

Well, I am studying Haskell Monads. When I read the Wikibook Category theory article, I found that the signature of monad morphisms looks pretty like tautologies in logic, but you need to convert M a to ~~A, here ~ is the logic negation.
return :: a -> M a -- Map to tautology A => ~~A, double negation introduction
(>>=) :: M a -> (a -> M b) -> M b -- Map to tautology ~~A => (A => ~~B) => ~~B
the other operations is also tautologies:
fmap :: (a -> b) -> M a -> M b -- Map to (A => B) -> (~~A => ~~B)
join :: M (M a) -> M a -- Map to ~~(~~A) => ~~A
It's also understood that according to the fact that the Curry-Howard correspondence of normal functional languages is the intuitive logic, not classical logic, so we cannot expect a tautology like ~~A => A can have a correspondence.
But I am thinking of something else. Why the Monad can only relate to a double negation? what is the correspondence of single negation? This lead me to the following class definition:
class Nomad n where
rfmap :: (a -> b) -> n b -> n a
dneg :: a -> n (n a)
return :: Nomad n => a -> n (n a)
return = dneg
(>>=) :: Nomad n => n (n a) -> (a -> n (n b)) -> n (n b)
x >>= f = rfmap dneg $ rfmap (rfmap f) x
Here I defined a concept called "Nomad", and it supports two operations (both related to a logic axiom in intuitive logic). Notice the name "rfmap" means the fact that its signature is similar to functor's fmap, but the order of a and b are reversed in result. Now I can re-define the Monad operations with them, with replace M a to n (n a).
So now let's go to the question part. The fact that Monad is a concept from category theory seems to mean that my "Nomad" is also a category theory concept. So what is it? Is it useful? Are there any papers or research results exists in this topic?

Double negation is a particular monad
data Void --no constructor because it is uninhabited
newtype DN a = DN {runDN :: (a -> Void) -> Void}
instance Monad DN where
return x = DN $ \f -> f x
m >>= k = DN $ \c -> runDN m (\a -> runDN (k a) c))
actually, this is an example of a more general monad
type DN = Cont Void
newtype Cont r a = Cont {runCont :: (a -> r) -> r}
which is the monad for continuation passing.
A concept like "monad" is not defined just by a signature, but also by some laws. So, here is a question, what should the laws for your construction be?
(a -> b) -> f b -> f a
is the signature of a method well known to category theory, the contravariant functor. It obeys essentially the same laws as functors (preserves (co)composition and identity). Actually, a contravariant functor is exactly a functor to the oppositite category. Since we are interested in "haskell functors" which are supposed to be endo-functors, we can see that a "haskell contravariant functor" is a functor Hask -> Hask_Op.
On the other hand, what about
a -> f (f a)
what laws should this have? I have a suggestion. In category theory, it is possible to map between Functors. Given two functors from F, G each from category C to category D, a natural transformation from F to G is a morphism in D
forall a. F a -> G a
that obeys certain coherence laws. You can do a lot with natural transformations, including using them to define a "functor category." But, the classic joke (due to Mac Lane) is that categories were invented to talk about functors, functors were invented to talk about natural transformations, and natural transformations were invented to talk about adjunctions.
A functor F : D -> C and a functor G : C -> D form an adjunction from C to D if there exist two natural transformations
unit : Id -> G . F
counit : F . G -> Id
This idea of an adjunction is a frequent way of understanding monads. Every adjunction gives rise to a monad in a totally natural way. That is, since the composition of these two functors is an endofunctor, and you have something akin to return (the unit), all you need is the join. But that is easy, join is just a function G . F . G . F -> G . F which you get by just using the counit "in the middle".
So, then with all this, what is it that you are looking for? Well
dneg :: a -> n (n a)
looks exactly like the unit of the adjunction of a contravariant functor with itself. Therefore, your Nomad type is likely (certainly if your use of it to construct a monad is correct) exactly the same as "a contravariant functor which is self adjoint." Searching for self adjoint functors will lead you back to Double-negation and Continuation passing...which is just where we started!
EDIT: Almost certainly some of the arrows above are backwards. The basic idea is right, though. You can work it out yourself using the citations below:
The best books on category theory are probably,
Steve Awodey, Category Theory
Sanders Mac Lane, Categories for the Working Mathematician
although numerous more approachable intro books exist including Benjamin Pierces book on Category Theory for computer scientists.
Video lectures online
The Category Theory lectures by Steve Awodey from the Oregon Programming Language Summer School
The Catsters short lectures on youtube, indexed here, pay particular attention to the videos on adjunctions
A number of papers explore the adjunction angle on the continuation monad, for example
Hayo Thielecke, Continuation Semantics and Self-adjointness
The search terms "self adjoint", "continuation", and "monad" are good. Also, a number of bloggers have written about these issues. If you google "where do monads come from" you get useful results like this one from the n category cafe, or this one from sigfpe. Also Sjoerd Vissche's link to the Comonad reader.

That would be a self-adjoint contravariant functor. rfmap provides the contravariant functor part, and dneg is the unit and counit of the adjunction.
Op r is an example, which creates the continuation monad. See the contravariant modules in http://hackage.haskell.org/package/adjunctions for some code.
You should read http://comonad.com/reader/2011/monads-from-comonads/, which is related and very interesting.

Related

`refold :: Functor s => (a -> s a, a) -> (s b -> b) -> b` as a morphism between universal types

Various recursion scheme boil down to specific instantiation of refold
refold :: Functor s => (s b -> b) -> (a -> s a) -> a -> b
refold f g = go where go a = f (fmap go (g a))
What is the meaningful interpretation of refold ?
The data type data Nu f = forall a. Nu (a -> f a) a and newtype Mu f = Mu {unMu :: forall b. (f b -> b) -> b} can be seen as the colimit and limit of the forget functor from coalgebras and algebras, and refold is a morphism between those, but does it shed light on refold ?
refold' :: forall s. Functor s => Nu s -> Mu s
refold' (Nu g (a :: a)) = Mu mu where
mu :: forall b. (s b -> b) -> b
mu f = go a where
go :: a -> b
go a = f (fmap go (g a))
I guess it depends what you mean by "meaningful interpretation".
If s is a base functor for a recursive data type and a corecursive codata type, like the following functor s ~ ListF e for the recursive list data type [e] (which, in Haskell, is also a corecursive stream codata type):
{-# LANGUAGE DeriveFunctor #-}
data ListF e b = Nil | Cons e b deriving (Show, Functor)
then an s-coalgebra of type a -> s a together with a starting seed a can generate a value of codata type [e] by unfolding from that seed, while an s-algebra of type s b -> b can consume a value of data type [e] by folding into a value of type b. The refold function just combines the operation of unfolding from a and folding into b, without actually creating an intermediate codata/data type.
For example, you can generate the (finite) codata stream [10,9..1] by unfolding from an Integer seed using the starting value / coalgebra pair (a,g) as follows:
a :: Integer
a = 10
g :: Integer -> (ListF Integer) Integer
g 0 = Nil
g n = Cons n (n-1)
and fold a list to calculate its Int length using the algebra:
f :: (ListF Integer) Int -> Int
f Nil = 0
f (Cons _ b) = 1 + b
The refold function just combines these operations:
main = print $ refold f g a
In this particular case, it calculates the length 10 of the stream/list [1..10] without actually creating any intermediate stream/list.
I guess the intuition is that if an operation can be imagined as an F-recursion applied to an F-corecursion for the same functor F, then it's a refold. Or, maybe more practically, if an algorithm has an internal recursive structure that matches the functor F, it can be expressed as a refold. The documentation for refold in recursion-schemes gives the example of quicksort having a recursive structure that matches a binary tree, though you've presumably already seen that example.
Note: What follows is wrong or at best imprecise, but I'll try to think a little more about it.
In practice, refold isn't only used as a morphism between universal data types, but if you have a final s-coalgebra for a codata type C associated with the functor s:
eatC :: C -> ListF Integer C
and an initial s-algebra for a data type D also associated with the functor s:
makeD :: ListF Integer D -> D
then refold makeD eatC should be a natural morphism from codata type C to data type D. That is, it should be the unique morphism satsifying:
fmap h . refold makeD eatC = refold makeD eatC . fmap h
I'm not sure that aspect is tremendously interesting...
A few remarks (which I imagine to be valid - don't hesitate to correct - I am not an expert on semantic) :
non termination allow to write anything, as #chi suggested.
take s the identity functor and refold reads refold :: (b -> b) -> (a -> a) -> a -> b which certainly looks like a paradox. So for any haskell type to be read "logically", we might need hidden, side conditions.
We don't even need recursion to encounter paradox / non termination
-- N. P. Mendler. Recursive types and type constraints in second-order lambda calculus. In LICS, pages 30–36. IEEE Computer Society, 1987
data T a = C (T a -> ())
p :: T a -> (T a ->() )
p (C f) = f
w :: T a -> ()
w x = (p x) x
initial algebra, like monads, and other concepts, happens on two level, one in the semantic of a language, and another explicitly in the program that we write. For instance the semantic of data ListInt = [] | Int * ListInt is an initial algebra. and, in haskell, also, semantically, a final coalgebra. This is "Mu = Nu" that you might hear,and this equation, which is paradoxical, happens is in the semantic of haskell. Mu here has nothing to do with data Mu f = ... The same happen when we write type ListInt = Fix ListIntF where data Fix f = Fix (f (Fix f)) we are emulating that semantic in our program, but this itself is subject to the semantic of Haskell (and indeed, this (initial algebra) is the same as the semantic Mu and Nu and equal to both because haskell equates them). In a way by writing data Mu f = ... and data Nu f = .. we are "stealing" (and obliged to do so) part of haskell semantic, as well as mixing it with our own (properly expressing the universal co-cone Nu f and the universal cone Mu f), attempting at providing an embedding for recursion (just like we do with HOAS where we steal bindings from Haskell). But we can't get away with no paradox, as we are obliged to steal that "Mu = Nu".
This leads to very useful and yet very "illogical" function like refold
By writing fold : Functor s => (f a -> a) -> Fix s -> a we are pretending the initial algebra of f always exists, which might translate to non-termination
Categorically we can view refold in two different ways. It's a bit of a mouthful but here we go :
refold can be seen as a proper function refold :: Functor s => (s b -> b) -> (a -> s a) -> a -> b to be detailed later
refold' can be seen as the carrier refold' :: forall s. Functor s => Nu s -> Mu s of an algebra in Twisted(Hask), whose objects are morphisms in Hask. So refold' is an object, not a morphism, of this category. Now every functor s on a category C (here Hask) induces a functor s' on Twisted(C) by applying to arrows. In the end the morphism in Twisted
`s' refold' -(out,in)-> refold'`
is the initial s' algebra, where out is the "final" coalgebra Nu s -> s (Nu s) and in is the "initial" algebra Mu s -> s (Mu s)
Now the action of the function refold is, given a coalgebra and an algebra (here in Hask but could be elsewhere), to return the unique morphism from the carrier of the coalgebra, followed by refold', followed by the unique morphism from the initial algebra. This is a proper function that comes from selecting the components of the universal (co)cone at the given components.
This explains why when we feed the final coalgebra out and the initial algebra in to refold, we get back refold' itself. the unique morphism to pre and post compose refold' with are the identities.
It's a bit hard to see what's what because we work in Hask where everything is a function. Some morphisms are really about the category we work in (could be other things than Hask) and some morphism really are functions would be even if we chose another category.
Because of non termination the solution to knowing what really is refold has to being truthful to the semantic of haskell, and use complete partial order (or to restrict s in some way).
So I imagine the real meaning of refold can be deduced from the real meaning of refold' which is just a initial algebra, with all the standard caveat coming from haskell semantic threaded.

What is the relationship between bind and join?

I got the impression that (>>=) (used by Haskell) and join (preferred by mathematicians) are "equal" since one can write one in terms of the other:
import Control.Monad (join)
join x = x >>= id
x >>= f = join (fmap f x)
Additionally every monad is a functor since bind can be used to replace fmap:
fmap f x = x >>= (return . f)
I have the following questions:
Is there a (non-recursive) definition of fmap in terms of join? (fmap f x = join $ fmap (return . f) x follows from the equations above but is recursive.)
Is "every monad is a functor" a conclusion when using bind (in the definition of a monad), but an assumption when using join?
Is bind more "powerful" than join? And what would "more powerful" mean?
A monad can be either defined in terms of:
return :: a -> m a
bind :: m a -> (a -> m b) -> m b
or alternatively in terms of:
return :: a -> m a
fmap :: (a -> b) -> m a -> m b
join :: m (m a) -> m a
To your questions:
No, we cannot define fmap in terms of join, since otherwise we could remove fmap from the second list above.
No, "every monad is a functor" is a statement about monads in general, regardless whether you define your specific monad in terms of bind or in terms of join and fmap. It is easier to understand the statement if you see the second definition, but that's it.
Yes, bind is more "powerful" than join. It is exactly as "powerful" as join and fmap combined, if you mean with "powerful" that it has the capacity to define a monad (always in combination with return).
For an intuition, see e.g. this answer – bind allows you to combine or chain strategies/plans/computations (that are in a context) together. As an example, let's use the Maybe context (or Maybe monad):
λ: let plusOne x = Just (x + 1)
λ: Just 3 >>= plusOne
Just 4
fmap also let's you chain computations in a context together, but at the cost of increasing the nesting with every step.[1]
λ: fmap plusOne (Just 3)
Just (Just 4)
That's why you need join: to squash two levels of nesting into one. Remember:
join :: m (m a) -> m a
Having only the squashing step doesn't get you very far. You need also fmap to have a monad – and return, which is Just in the example above.
[1]: fmap and (>>=) don't take their two arguments in the same order, but don't let that confuse you.
Is there a [definition] of fmap in terms of join?
No, there isn't. That can be demonstrated by attempting to do it. Suppose we are given an arbitrary type constructor T, and functions:
returnT :: a -> T a
joinT :: T (T a) -> T a
From this data alone, we want to define:
fmapT :: (a -> b) -> T a -> T b
So let's sketch it:
fmapT :: (a -> b) -> T a -> T b
fmapT f ta = tb
where
tb = undefined -- tb :: T b
We need to get a value of type T b somehow. ta :: T a on its own won't do, so we need functions that produce T b values. The only two candidates are joinT and returnT. joinT doesn't help:
fmapT :: (a -> b) -> T a -> T b
fmapT f ta = joinT ttb
where
ttb = undefined -- ttb :: T (T b)
It just kicks the can down the road, as needing a T (T b) value under these circumstances is no improvement.
We might try returnT instead:
fmapT :: (a -> b) -> T a -> T b
fmapT f ta = returnT b
where
b = undefined -- b :: b
Now we need a b value. The only thing that can give us one is f:
fmapT :: (a -> b) -> T a -> T b
fmapT f ta = returnT (f a)
where
a = undefined -- a :: a
And now we are stuck: nothing can give us an a. We have exhausted all available possibilities, so fmapT cannot be defined in such terms.
A digression: it wouldn't suffice to cheat by using a function like this:
extractT :: T a -> a
With an extractT, we might try a = extractT ta, leading to:
fmapT :: (a -> b) -> T a -> T b
fmapT f ta = returnT (f (extractT ta))
It is not enough, however, for fmapT to have the right type: it must also follow the functor laws. In particular, fmapT id = id should hold. With this definition, fmapT id is returnT . extractT, which, in general, is not id (most functors which are instances of both Monad and Comonad serve as examples).
Is "every monad is a functor" a conclusion when using bind (in the definition of a monad), but an assumption when using join?
"Every monad is a functor" is an assumption, or, more precisely, part of the definition of monad. To pick an arbitrary illustration, here is Emily Riehl, Category Theory In Context, p. 154:
Definition 5.1.1. A monad on a category C consists of
an endofunctor T : C → C,
a unit natural transformation η : 1C ⇒ T, and
a multiplication natural transformation μ :T2 ⇒ T,
so that the following diagrams commute in CC: [diagrams of the monad laws]
A monad, therefore, involves an endofunctor by definition. For a Haskell type constructor T that instantiates Monad, the object mapping of that endofunctor is T itself, and the morphism mapping is its fmap. That T will be a Functor instance, and therefore will have an fmap, is, in contemporary Haskell, guaranteed by Applicative (and, by extension, Functor) being a superclass of Monad.
Is that the whole story, though? As far as Haskell is concerned. we know that liftM exists, and also that in a not-so-distant past Functor was not a superclass of Monad. Are those two facts mere Haskellisms? Not quite. In the classic paper Notions of computation and monads, Eugenio Moggi unearths the following definition (p. 3):
Definition 1.2 ([Man76]) A Kleisli triple over a category C is a triple (T, η, _*), where T : Obj(C) → Obj(C), ηA : A → T A for A ∈ Obj(C), f* : T A → T B for f : A → T B and the following equations hold:
ηA* = idT A
ηA; f* = f   for   f : A → T B
f*; g* = (f; g*)*   for   f : A → T B   and   g : B → T C
The important detail here is that T is presented as merely an object mapping in the category C, and not as an endofunctor in C. Working in the Hask category, that amounts to taking a type constructor T without presupposing it is a Functor instance. In code, we might write that as:
class KleisliTriple t where
return :: a -> t a
(=<<) :: (a -> t b) -> t a -> t b
-- (return =<<) = id
-- (f =<<) . return = f
-- (g =<<) . (f =<<) = ((g =<<) . f =<<)
Flipped bind aside, that is the pre-AMP definition of Monad in Haskell. Unsurprisingly, Moggi's paper doesn't take long to show that "there is a one-to-one correspondence between Kleisli triples and monads" (p. 5), establishing along the way that T can be extended to an endofunctor (in Haskell, that step amounts to defining the morphism mapping liftM f m = return . f =<< m, and then showing it follows the functor laws).
All in all, if you write lawful definitions of return and (>>=) without presupposing fmap, you indeed get a lawful implementation of Functor as a consequence. "There is a one-to-one correspondence between Kleisli triples and monads" is a consequence of the definition of Kleisli triple, while "a monad involves an endofunctor" is part of the definition of monad. It is tempting to consider whether it would be more accurate to describe what Haskellers did when writing Monad instances as "setting up a Klesili triple" rather than "setting up a monad", but I will refrain out of fear of getting mired down terminological pedantry -- and in any case, now that Functor is a superclass of Monad there is no practical reason to worry about that.
Is bind more "powerful" than join? And what would "more powerful" mean?
Trick question!
Taken at face value, the answer would be yes, to the extent that, together with return, (>>=) makes it possible to implement fmap (via liftM, as noted above), while join doesn't. However, I don't feel it is worthwhile to insist on this distinction. Why so? Because of the monad laws. Just like it doesn't make sense to talk about a lawful (>>=) without presupposing return, it doesn't make sense to talk about a lawful join without pressuposing return and fmap.
One might get the impression that I am giving too much weight to the laws by using them to tie Monad and Functor in this way. It is true that there are cases of laws that involve two classes, and that only apply to types which instantiate them both. Foldable provides a good example of that: we can find the following law in the Traversable documentation:
The superclass instances should satisfy the following: [...]
In the Foldable instance, foldMap should be equivalent to traversal with a constant applicative functor (foldMapDefault).
That this specific law doesn't always apply is not a problem, because we don't need it to characterise what Foldable is (alternatives include "a Foldable is a container from which we can extract some sequence of elements", and "a Foldable is a container that can be converted to the free monoid on its element type"). With the monad laws, though, it isn't like that: the meaning of the class is inextricably bound to all three of the monad laws.

How does lifting (in a functional programming context) relate to category theory?

Looking at the Haskell documentation, lifting seems to be basically a generalization of fmap, allowing for the mapping of functions with more than one argument.
The Wikipedia article on lifting gives a different view however, defining a "lift" in terms of a morphism in a category, and how it relates to the other objects and morphisms in the category (I won't give the details here). I suppose that that could be relevant to the Haskell situation if we are considering Cat (the category of categories, thus making our morphisms functors), but I can't see how this category-theoretic notion of a lift relates the the one in Haskell based on the linked article, if it does at all.
If the two concepts aren't really related, and just have a similar name, are the lifts (category theory) used in Haskell at all?
Lifts, and the dual notion of extensions, are absolutely used in Haskell, perhaps most prominently in the guise of comonadic extend and monadic bind. (Confusingly, extend is a lift, not an extension.) A comonad w's extend lets us take a function w a -> b and lift it along extract :: w b -> b to get a map w a -> w b. In ASCII art, given the diagram
w b
|
V
w a ---> b
where the vertical arrow is extract, extend gives us a diagonal arrow (making the diagram commute):
-> w b
/ |
/ V
w a ---> b
More familiar to most Haskellers is the dual notion of bind (>>=) for a monad m. Given a function a -> m b and return :: a -> m a, we can "extend" our function along return to get a function m a -> m b. In ASCII art:
a ---> m b
|
V
m a
gives us
a ---> m b
| __A
V /
m a
(That A is an arrowhead!)
So yes, extend could have been called lift, and bind could have been called extend. As for Haskell's lifts, I have no idea why they're called that!
EDIT: Actually, I think that again, Haskell's lifts are actually extensions. If f is applicative, and we have a function a -> b -> c, we can compose this function with pure :: c -> f c to get a function a -> b -> f c. Uncurrying, this is the same as a function (a, b) -> f c. Now we can also hit (a, b) with pure to get a function (a, b) -> f (a, b). Now, by fmaping fst and snd, we get a functions f (a, b) -> f a and f (a, b) -> f b, which we can combine to get a function f (a, b) -> (f a, f b). Composing with our pure from before gives (a, b) -> (f a, f b). Phew! So to recap, we have the ASCII art diagram
(a, b) ---> f c
|
V
(f a, f b)
Now liftA2 gives us a function (f a, f b) -> f c, which I won't draw because I'm sick of making terrible diagrams. But the point is, the diagram commutes, so liftA2 actually gives us an extension of the horizontal arrow along the vertical one.
"Lifting" comes up many times in functional programming, not only in fmap but in many other contexts. Examples of "liftings" include:
fmap :: (a -> b) -> F a -> F b where F is a functor
cmap :: (b -> a) -> F a -> F b where F is a contrafunctor
bind :: (a -> M b) -> M a -> M b where M is a monad
ap :: F (a -> b) -> F a -> F b where F is an applicative functor
point :: (_ -> a) -> _ -> F a where F is a pointed functor
filtMap :: (a -> Maybe b) -> F a -> F b where F is a filterable functor
extend :: (M a -> b) -> M a -> M b where M is a comonad
Other examples include applicative contrafunctor, filterable contrafunctor, and co-pointed functor.
All these type signatures are similar in one way: they map one kind of function between a and b into another kind of function between a and b.
In these different cases, the function types are not simply a -> b but have some kind of "twisted" types: e.g. a -> M b or F (a -> b) or M a -> b or F a -> F b and so on. However, each time the laws are very similar: twisted function types need to have identity and composition laws, and twisted composition needs to be associative.
For example, for applicative functors, we need to be able to compose functions of type F (a -> b). So we need to define a special "twisted" identity function (pure id :: F (a -> a) ) and a "twisted" composition operation, call it apcomp, with type signature F (a -> b) -> F (b -> c) -> F (a -> c). This operation needs to have identity and associativity laws. The ap operation needs to have identity and composition laws ("twisted identity maps to twisted identity" and "twisted composition maps to twisted composition").
Once we go through all these examples and derive the laws, we can prove that the laws turn out to be the same in all cases, if we formulate the laws via the "twisted" operations.
This is because we can formulate all these operations as functors in the sense of category theory. For example, for the applicative functor, we define two categories: the F-applicative category (objects a, b, ..., morphisms F(a -> b)) and the F-lifted category (objects F a, F b, ..., morphisms F a -> F b). A functor between these two categories requires us to have a lifting of morphisms, ap :: F(a -> b) -> F a -> F b. The laws of ap are completely equivalent to the standard laws of that functor.
Similar arguments hold for other typeclasses. We need to define categories, morphisms, composition operations, identity morphisms, and functors in each case. Once we verify that the laws hold, we will see that each of these typeclasses has an associated pair of categories and a functor between them, such that the laws of the typeclass are equivalent to the laws of these categories and the functor.
What have we gained? We have formulated the laws of many typeclasses in the same way (as the laws of categories and functors). This is a great economy of thought: we don't need to memorize all these laws each time; we can just memorize which categories and which functors need to be written down for each typeclass, as long as the methods of the typeclass can be reduced to some kind of "twisted lifting".
In this way, we can say that "liftings" are important and provide an application of category theory in functional programming.
I have made a presentation about this, https://www.youtube.com/watch?v=Zau8CxsfxOo and I'm writing a new free book where all derivations will be shown. https://github.com/winitzki/sofp

What are free monads?

I've seen the term Free Monad pop up every now and then for some time, but everyone just seems to use/discuss them without giving an explanation of what they are. So: what are free monads? (I'd say I'm familiar with monads and the Haskell basics, but have only a very rough knowledge of category theory.)
Here's an even simpler answer: A Monad is something that "computes" when monadic context is collapsed by join :: m (m a) -> m a (recalling that >>= can be defined as x >>= y = join (fmap y x)). This is how Monads carry context through a sequential chain of computations: because at each point in the series, the context from the previous call is collapsed with the next.
A free monad satisfies all the Monad laws, but does not do any collapsing (i.e., computation). It just builds up a nested series of contexts. The user who creates such a free monadic value is responsible for doing something with those nested contexts, so that the meaning of such a composition can be deferred until after the monadic value has been created.
Edward Kmett's answer is obviously great. But, it is a bit technical. Here is a perhaps more accessible explanation.
Free monads are just a general way of turning functors into monads. That is, given any functor f Free f is a monad. This would not be very useful, except you get a pair of functions
liftFree :: Functor f => f a -> Free f a
foldFree :: Functor f => (f r -> r) -> Free f r -> r
the first of these lets you "get into" your monad, and the second one gives you a way to "get out" of it.
More generally, if X is a Y with some extra stuff P, then a "free X" is a a way of getting from a Y to an X without gaining anything extra.
Examples: a monoid (X) is a set (Y) with extra structure (P) that basically says it has an operation (you can think of addition) and some identity (like zero).
So
class Monoid m where
mempty :: m
mappend :: m -> m -> m
Now, we all know lists
data [a] = [] | a : [a]
Well, given any type t we know that [t] is a monoid
instance Monoid [t] where
mempty = []
mappend = (++)
and so lists are the "free monoid" over sets (or in Haskell types).
Okay, so free monads are the same idea. We take a functor, and give back a monad. In fact, since monads can be seen as monoids in the category of endofunctors, the definition of a list
data [a] = [] | a : [a]
looks a lot like the definition of free monads
data Free f a = Pure a | Roll (f (Free f a))
and the Monad instance has a similarity to the Monoid instance for lists
--it needs to be a functor
instance Functor f => Functor (Free f) where
fmap f (Pure a) = Pure (f a)
fmap f (Roll x) = Roll (fmap (fmap f) x)
--this is the same thing as (++) basically
concatFree :: Functor f => Free f (Free f a) -> Free f a
concatFree (Pure x) = x
concatFree (Roll y) = Roll (fmap concatFree y)
instance Functor f => Monad (Free f) where
return = Pure -- just like []
x >>= f = concatFree (fmap f x) --this is the standard concatMap definition of bind
now, we get our two operations
-- this is essentially the same as \x -> [x]
liftFree :: Functor f => f a -> Free f a
liftFree x = Roll (fmap Pure x)
-- this is essentially the same as folding a list
foldFree :: Functor f => (f r -> r) -> Free f r -> r
foldFree _ (Pure a) = a
foldFree f (Roll x) = f (fmap (foldFree f) x)
A free foo happens to be the simplest thing that satisfies all of the 'foo' laws. That is to say it satisfies exactly the laws necessary to be a foo and nothing extra.
A forgetful functor is one that "forgets" part of the structure as it goes from one category to another.
Given functors F : D -> C, and G : C -> D, we say F -| G, F is left adjoint to G, or G is right adjoint to F whenever forall a, b: F a -> b is isomorphic to a -> G b, where the arrows come from the appropriate categories.
Formally, a free functor is left adjoint to a forgetful functor.
The Free Monoid
Let us start with a simpler example, the free monoid.
Take a monoid, which is defined by some carrier set T, a binary function to mash a pair of elements together f :: T → T → T, and a unit :: T, such that you have an associative law, and an identity law: f(unit,x) = x = f(x,unit).
You can make a functor U from the category of monoids (where arrows are monoid homomorphisms, that is, they ensure they map unit to unit on the other monoid, and that you can compose before or after mapping to the other monoid without changing meaning) to the category of sets (where arrows are just function arrows) that 'forgets' about the operation and unit, and just gives you the carrier set.
Then, you can define a functor F from the category of sets back to the category of monoids that is left adjoint to this functor. That functor is the functor that maps a set a to the monoid [a], where unit = [], and mappend = (++).
So to review our example so far, in pseudo-Haskell:
U : Mon → Set -- is our forgetful functor
U (a,mappend,mempty) = a
F : Set → Mon -- is our free functor
F a = ([a],(++),[])
Then to show F is free, we need to demonstrate that it is left adjoint to U, a forgetful functor, that is, as we mentioned above, we need to show that
F a → b is isomorphic to a → U b
now, remember the target of F is in the category Mon of monoids, where arrows are monoid homomorphisms, so we need a to show that a monoid homomorphism from [a] → b can be described precisely by a function from a → b.
In Haskell, we call the side of this that lives in Set (er, Hask, the category of Haskell types that we pretend is Set), just foldMap, which when specialized from Data.Foldable to Lists has type Monoid m => (a → m) → [a] → m.
There are consequences that follow from this being an adjunction. Notably that if you forget then build up with free, then forget again, its just like you forgot once, and we can use this to build up the monadic join. since UFUF ~ U(FUF) ~ UF, and we can pass in the identity monoid homomorphism from [a] to [a] through the isomorphism that defines our adjunction,get that a list isomorphism from [a] → [a] is a function of type a -> [a], and this is just return for lists.
You can compose all of this more directly by describing a list in these terms with:
newtype List a = List (forall b. Monoid b => (a -> b) -> b)
The Free Monad
So what is a Free Monad?
Well, we do the same thing we did before, we start with a forgetful functor U from the category of monads where arrows are monad homomorphisms to a category of endofunctors where the arrows are natural transformations, and we look for a functor that is left adjoint to that.
So, how does this relate to the notion of a free monad as it is usually used?
Knowing that something is a free monad, Free f, tells you that giving a monad homomorphism from Free f -> m, is the same thing (isomorphic to) as giving a natural transformation (a functor homomorphism) from f -> m. Remember F a -> b must be isomorphic to a -> U b for F to be left adjoint to U. U here mapped monads to functors.
F is at least isomorphic to the Free type I use in my free package on hackage.
We could also construct it in tighter analogy to the code above for the free list, by defining
class Algebra f x where
phi :: f x -> x
newtype Free f a = Free (forall x. Algebra f x => (a -> x) -> x)
Cofree Comonads
We can construct something similar, by looking at the right adjoint to a forgetful functor assuming it exists. A cofree functor is simply /right adjoint/ to a forgetful functor, and by symmetry, knowing something is a cofree comonad is the same as knowing that giving a comonad homomorphism from w -> Cofree f is the same thing as giving a natural transformation from w -> f.
The Free Monad (data structure) is to the Monad (class) like the List (data structure) to the Monoid (class): It is the trivial implementation, where you can decide afterwards how the content will be combined.
You probably know what a Monad is and that each Monad needs a specific (Monad-law abiding) implementation of either fmap + join + return or bind + return.
Let us assume you have a Functor (an implementation of fmap) but the rest depends on values and choices made at run-time, which means that you want to be able to use the Monad properties but want to choose the Monad-functions afterwards.
That can be done using the Free Monad (data structure), which wraps the Functor (type) in such a way so that the join is rather a stacking of those functors than a reduction.
The real return and join you want to use, can now be given as parameters to the reduction function foldFree:
foldFree :: Functor f => (a -> b) -> (f b -> b) -> Free f a -> b
foldFree return join :: Monad m => Free m a -> m a
To explain the types, we can replace Functor f with Monad m and b with (m a):
foldFree :: Monad m => (a -> (m a)) -> (m (m a) -> (m a)) -> Free m a -> (m a)
A Haskell free monad is a list of functors. Compare:
data List a = Nil | Cons a (List a )
data Free f r = Pure r | Free (f (Free f r))
Pure is analogous to Nil and Free is analogous to Cons. A free monad stores a list of functors instead of a list of values. Technically, you could implement free monads using a different data type, but any implementation should be isomorphic to the above one.
You use free monads whenever you need an abstract syntax tree. The base functor of the free monad is the shape of each step of the syntax tree.
My post, which somebody already linked, gives several examples of how to build abstract syntax trees with free monads
I think a simple concrete example will help. Suppose we have a functor
data F a = One a | Two a a | Two' a a | Three Int a a a
with the obvious fmap. Then Free F a is the type of trees whose leaves have type a and whose nodes are tagged with One, Two, Two' and Three. One-nodes have one child, Two- and Two'-nodes have two children and Three-nodes have three and are also tagged with an Int.
Free F is a monad. return maps x to the tree that is just a leaf with value x. t >>= f looks at each of the leaves and replaces them with trees. When the leaf has value y it replaces that leaf with the tree f y.
A diagram makes this clearer, but I don't have the facilities for easily drawing one!
Trying to provide a “bridge” answer between the auper-simple answers here and the full answer.
So “free monads” build a “monad” out of any “functor”, so let’s take these in order.
Functors in detail
Some things are type-level adjectives, meaning they take a type-noun like “integers” and give you back a different type-noun like “lists of integers” or “pairs of strings with integers” or even “functions making strings out of integers.” To denote an arbitrary adjective let me use the stand-in word “blue”.
But then we notice a pattern that some of these adjectives are input-ish or output-ish in the noun they modify. For example “functions making strings out of __” is input-ish, “functions turning strings into __” is output-ish. The rule here involves me having a function X → Y, some adjective “blue” is outputtish, or a functor, if I can use such a function to transform a blue X into a blue Y. Think of “a firehose spraying Xs” and then you screw on this X → Y function and now your firehose sprays Ys. Or it is inputtish or a contravariant if it is the opposite, a vacuum cleaner sucking up Ys and when I screw this on I get a vacuum sucking up Xs. Some things are neither outputtish nor inputtish. Things that are both it turns out are phantom: they have absolutely nothing to do with the nouns that they describe, in the sense that you can define a function “coerce” which takes a blue X and makes a blue Y, but *without knowing the details of the types X or Y,” not even requiring a function between them.
So “lists of __” is outputtish, you can map an X → Y over a list of Xs to get a list of Ys. Similarly “a pair of a string and a __” is outputtish. Meanwhile “a function from __ to itself” is neither outputtish nor inputtish,” while “a string and a list of exactly zero __s” could be the “phantom” case maybe.
But yeah, that's all there is to functors in programming, they are just things that are mappable. If something is a functor it is a functor uniquely, there is at most only one way to generically map a function over a data structure.
Monads
A monad is a functor that in addition is both
Universally applicable, given any X, I can construct a blue X,
Can be repeated without changing the meaning much. So a blue-blue X is in some sense the same as just a blue X.
So what this means is that there is a canonical function collapsing any blue-blue __ to just a blue __. (And we of course add laws to make everything sane: if one of the layers of blue came from the universal application, then we want to just erase that universal application; in addition if we flatten a blue-blue-blue X down to a blue X, it should not make a difference whether we collapse the first two blues first or the second two first.)
The first example is “a nullable __”. So if I give you a nullable nullable int, in some sense I have not given you much more than a nullable int. Or “a list of lists of ints,” if the point is just to have 0 or more of them, then “a list of ints” works just fine and the proper collapse is concatenating all of the lists together into one super list.
Monads became big in Haskell because Haskell needed an approach to do things in the real world without violating its mathematically pure world where nothing really happens. The solution was to add is sort of watered down form of metaprogramming where we introduce an adjective of “a program which produces a __.” So how do I fetch the current date? Well, Haskell can't do it directly without unsafePerformIO but it will let you describe and compose the program which produces the current date. In this vision, you are supposed to describe a program which produces nothing called “Main.main,” and the compiler is supposed to take your description and hand you this program as a binary executable for you to run whenever you want.
Anyway “a program which produces a program which produce an int” doesn't buy you much over “a program which produces an int” so this turns out to be a monad.
Free Monads
Unlike functors, monads aren't uniquely monads. There is not just one monad instance for a given functor. So for example “a pair of an int and a __”, what are you doing with that int? You could add it, you could multiply it. If you make it a nullable int, you could keep the minimum non-null one or the maximum non-null one, or the leftmost non-null one or the rightmost non-null one.
The free monad for a given functor is the "boringest” construction, it is just “A free blue X is a bluen X for any n = 0, 1, 2, ...”.
It is universal because a blue⁰ X is just an X. And a free blue free blue X is a bluem bluen X which is just a bluem+n X. It implements “collapse” therefore by not implementing collapse at all, internally the blues are nested arbitrarily.
This also means you can defer exactly which monad you are choosing until a later date, later you can define a function which reduces a blue-blue X to a blue X and collapse all of these to blue0,1 X and then another function from X to blue X gives you blue1 X throughout.

Computation Constructs (Monads, Arrows, etc.)

I have become rather interested in how computation is modeled in Haskell. Several resources have described monads as "composable computation" and arrows as "abstract views of computation". I've never seen monoids, functors or applicative functors described in this way. It seems that they lack the necessary structure.
I find that idea interesting and wonder if there are any other constructs that do something similar. If so, what are some resources that I can use to acquaint myself with them? Are there any packages on Hackage that might come in handy?
Note: This question is similar to
Monads vs. Arrows and https://stackoverflow.com/questions/2395715/resources-for-learning-monads-functors-monoids-arrows-etc, but I am looking for constructs beyond funtors, applicative functors, monads, and arrows.
Edit: I concede that applicative functors should be considered "computational constructs", but I'm really looking for something I haven't come across yet. This includes applicative functors, monads and arrows.
Arrows are generalized by Categories, and so by the Category typeclass.
class Category f where
(.) :: f a b -> f b c -> f a c
id :: f a a
The Arrow typeclass definition has Category as a superclass. Categories (in the haskell sense) generalize functions (you can compose them but not apply them) and so are definitely a "model of computation". Arrow provides a Category with additional structure for working with tuples. So, while Category mirrors something about Haskell's function space, Arrow extends that to something about product types.
Every Monad gives rise to something called a "Kleisli Category" and this construction gives you instances of ArrowApply. You can build a Monad out of any ArrowApply such that going full circle doesn't change your behavior, so in some deep sense Monad and ArrowApply are the same thing.
newtype Kleisli m a b = Kleisli { runKleisli :: a -> m b }
instance Monad m => Category (Kleisli m) where
id = Kleisli return
(Kleisli f) . (Kleisli g) = Kleisli (\b -> g b >>= f)
instance Monad m => Arrow (Kleisli m) where
arr f = Kleisli (return . f)
first (Kleisli f) = Kleisli (\ ~(b,d) -> f b >>= \c -> return (c,d))
second (Kleisli f) = Kleisli (\ ~(d,b) -> f b >>= \c -> return (d,c))
Actually every Arrow gives rise to an Applicative (universally quantified to get the kinds right) in addition to the Category superclass, and I believe the combination of the appropriate Category and Applicative is enough to reconstruct your Arrow.
So, these structures are deeply connected.
Warning: wishy-washy commentary ahead. One central difference between the Functor/Applicative/Monad way of thinking and the Category/Arrow way of thinking is that while Functor and its ilk are generalizations at the level of object (types in Haskell), Category/Arrow are generelazation of the notion of morphism (functions in Haskell). My belief is that thinking at the level of generalized morphism involves a higher level of abstraction than thinking at the level of generalized objects. Sometimes that is a good thing, other times it is not. On the other-hand, despite the fact that Arrows have a categorical basis, and no one in math thinks Applicative is interesting, it is my understanding that Applicative is generally better understood than Arrow.
Basically you can think of "Category < Arrow < ArrowApply" and "Functor < Applicative < Monad" such that "Category ~ Functor", "Arrow ~ Applicative" and "ArrowApply ~ Monad".
More Concrete Below:
As for other structures to model computation: one can often reverse the direction of the "arrows" (just meaning morphisms here) in categorical constructions to get the "dual" or "co-construction". So, if a monad is defined as
class Functor m => Monad m where
return :: a -> m a
join :: m (m a) -> m a
(okay, I know that isn't how Haskell defines things, but ma >>= f = join $ fmap f ma and join x = x >>= id so it just as well could be)
then the comonad is
class Functor m => Comonad m where
extract :: m a -> a -- this is co-return
duplicate :: m a -> m (m a) -- this is co-join
This thing turns out to be pretty common also. It turns out that Comonad is the basic underlying structure of cellular automata. For completness, I should point out that Edward Kmett's Control.Comonad puts duplicate in a class between functor and Comonad for "Extendable Functors" because you can also define
extend :: (m a -> b) -> m a -> m b -- Looks familiar? this is just the dual of >>=
extend f = fmap f . duplicate
--this is enough
duplicate = extend id
It turns out that all Monads are also "Extendable"
monadDuplicate :: Monad m => m a -> m (m a)
monadDuplicate = return
while all Comonads are "Joinable"
comonadJoin :: Comonad m => m (m a) -> m a
comonadJoin = extract
so these structures are very close together.
All Monads are Arrows (Monad is isomorphic to ArrowApply). In a different way, all Monads are instances of Applicative, where <*> is Control.Monad.ap and *> is >>. Applicative is weaker because it does not guarantee the >>= operation. Thus Applicative captures computations that do not examine previous results and branch on values. In retrospect much monadic code is actually applicative, and with a clean rewrite this would happen.
Extending monads, with recent Constraint kinds in GHC 7.4.1 there can now be nicer designs for restricted monads. And there are also people looking at parameterized monads, and of course I include a link to something by Oleg.
In libraries these structures give rise to different type of computations.
For example Applicatives can be used to implement static effects. With that I mean effects, which are defined at forehand. For example when implementing a state machine, rejecting or accepting an input state. They can't be used to manipulate their internal structure in terms of their input.
The type says it all:
<*> :: f (a -> b) -> f a -> f b
It is easy to reason, the structure of f cannot be depend om the input of a. Because a cannot reach f on the type level.
Monads can be used for dynamic effects. This also can be reasoned from the type signature:
>>= :: m a -> (a -> m b) -> m b
How can you see this? Because a is on the same "level" as m. Mathematically it is a two stage process. Bind is a composition of two function: fmap and join. First we use fmap together with the monadic action to create a new structure embedded in the old one:
fmap :: (a -> b) -> m a -> m b
f :: (a -> m b)
m :: m a
fmap f :: m a -> m (m b)
fmap f m :: m (m b)
Fmap can create a new structure, based on the input value. Then we collapse the structure with join, thus we are able to manipulate the structure from within the monadic computation in a way that depends on the input:
join :: m (m a) -> m a
join (fmap f m) :: m b
Many monads are easier to implement with join:
(>>=) = join . fmap
This is possible with monads:
addCounter :: Int -> m Int ()
But not with applicatives, but applicatives (and any monad) can do things like:
addOne :: m Int ()
Arrows give more control over the input and the output types, but for me they really feel similar to applicatives. Maybe I am wrong about that.

Resources