Following a minimal example of an observation (that kind of astonished me):
type Vector = V of float*float
// complete unfolding of type is OK
let projX (V (a,_)) = a
// also works
let projX' x =
match x with
| V (a, _) -> a
// BUT:
// partial unfolding is not Ok
let projX'' (V x) = fst x
// consequently also doesn't work
let projX''' x =
match x with
| V y -> fst y
What is the reason that makes it impossible to match against a partially deconstructed type?
Some partial deconstructions seem to be ok:
// Works
let f (x,y) = fst y
EDIT:
Ok, I now understand the "technical" reason of the behavior described (Thanks for your answers & comments). However, I think that language wise, this behavior feels a bit "unnatural" compared to rest of the language:
"Algebraically", to me, it seems strange to distinguish a type "t" from the type "(t)". Brackets (in this context) are used for giving precedence like e.g. in "(t * s) * r" vs "t * (s * r)". Also fsi answers accordingly, whether I send
type Vector = (int * int)
or
type Vector = int * int
to fsi, the answer is always
type Vector = int * int
Given those observations, one concludes that "int * int" and "(int * int)" denote exactly the same types and thus that all occurrences of one could in any piece of code be replaced with the other (ref. transparency)... which as we have seen is not true.
Further it seems significant that in order to explain the behavior at hand, we had to resort to talk about "how some code looks like after compilation" rather than about semantic properties of the language which imo indicates that there are some "tensions" between language semantics an what the compiler actually does.
In F#
type Vector = V of float*float
is just a degenerated union (you can see that by hovering it in Visual Studio), so it's equivalent to:
type Vector =
| V of float*float
The part after of creates two anonymous fields (as described in F# reference) and a constructor accepting two parameters of type float.
If you define
type Vector2 =
| V2 of (float*float)
there's only one anonymous field which is a tuple of floats and a constructor with a single parameter. As it was pointed out in the comment, you can use Vector2 to do desired pattern matching.
After all of that, it may seem illogical that following code works:
let argsTuple = (1., 1.)
let v1 = V argsTuple
However, if you take into account that there's a hidden pattern matching, everything should be clear.
EDIT:
F# language spec (p 122) states clearly that parenthesis matter in union definitions:
Parentheses are significant in union definitions. Thus, the following two definitions differ:
type CType = C of int * int
type CType = C of (int * int)
The lack of parentheses in the first example indicates that the union case takes two arguments. The parentheses
in the second example indicate that the union case takes one argument that is a first-class tuple value.
I think that such behavior is consistent with the fact that you can define more complex patterns at the definition of a union, e.g.:
type Move =
| M of (int * int) * (int * int)
Being able to use union with multiple arguments also makes much sense, especially in interop situation, when using tuples is cumbersome.
The other thing that you used:
type Vector = int * int
is a type abbreviation which simply gives a name to a certain type. Placing parenthesis around int * int does not make a difference because those parenthesis will be treated as grouping parenthesis.
Related
A function can be a highly nested structure:
function a(x) {
return b(c(x), d(e(f(x), g())))
}
First, wondering if a function has an instance. That is, the evaluation of the function being the instance of the function. In that sense, the type is the function, and the instance is the evaluation of it. If it can be, then how to model a function as a type (in some type-theory oriented language like Haskell or Coq).
It's almost like:
type a {
field: x
constructor b {
constructor c {
parameter: x
},
...
}
}
But I'm not sure if I'm not on the right track. I know you can say a function has a [return] type. But I'm wondering if a function can be considered a type, and if so, how to model it as a type in a type-theory-oriented language, where it models the actual implementation of the function.
I think the problem is that types based directly on the implementation (let's call them "i-types") don't seem very useful, and we already have good ways of modelling them (called "programs" -- ha ha).
In your specific example, the full i-type of your function, namely:
type a {
field: x
constructor b {
constructor c {
parameter: x
},
constructor d {
constructor e {
constructor f {
parameter: x
}
constructor g {
}
}
}
}
is just a verbose, alternative syntax for the implementation itself. That is, we could write this i-type (in a Haskell-like syntax) as:
itype a :: a x = b (c x) (d (e (f x) g))
On the other hand, we could convert your function implementation to Haskell term-level syntax directly to write it as:
a x = b (c x) (d (e (f x) g))
and the i-type and the implementation are exactly the same thing.
How would you use these i-types? The compiler might use them by deriving argument and return types to type-check the program. (Fortunately, there are well known algorithms, such as Algorithm W, for simultaneously deriving and type-checking argument and return types from i-types of this sort.) Programmers probably wouldn't use i-types directly -- they're too complicated to use for refactoring or reasoning about program behavior. They'd probably want to look at the types derived by the compiler for the arguments and return type.
In particular, "modelling" these i-types at the type level in Haskell doesn't seem productive. Haskell can already model them at the term level. Just write your i-types as a Haskell program:
a x = b (c x) (d (e (f x) g))
b s t = sqrt $ fromIntegral $ length (s ++ t)
c = show
d = reverse
e c ds = show (sum ds + fromIntegral (ord c))
f n = if even n then 'E' else 'O'
g = [1.5..5.5]
and don't run it. Congratulations, you've successfully modelled these i-types! You can even use GHCi to query derived argument and return types:
> :t a
a :: Floating a => Integer -> a -- "a" takes an Integer and returns a float
>
Now, you are perhaps imagining that there are situations where the implementation and i-type would diverge, maybe when you start introducing literal values. For example, maybe you feel like the function f above:
f n = if even n then 'E' else 'O'
should be assigned a type something like the following, that doesn't depend on the specific literal values:
type f {
field: n
if_then_else {
constructor even { -- predicate
parameter: n
}
literal Char -- then-branch
literal Char -- else-branch
}
Again, though, you'd be better off defining an arbitrary term-level Char, like:
someChar :: Char
someChar = undefined
and modeling this i-type at the term-level:
f n = if even n then someChar else someChar
Again, as long as you don't run the program, you've successfully modelled the i-type of f, can query its argument and return types, type-check it as part of a bigger program, etc.
I'm not clear exactly what you are aiming at, so I'll try to point at some related terms that you might want to read about.
A function has not only a return type, but a type that describes its arguments as well. So the (Haskell) type of f reads "f takes an Int and a Float, and returns a List of Floats."
f :: Int -> Float -> [Float]
f i x = replicate i x
Types can also describe much more of the specification of a function. Here, we might want the type to spell out that the length of the list will be the same as the first argument, or that every element of the list will be the same as the second argument. Length-indexed lists (often called Vectors) are a common first example of Dependent Types.
You might also be interested in functions that take types as arguments, and return types. These are sometimes called "type-level functions". In Coq or Idris, they can be defined the same way as more familiar functions. In Haskell, we usually implement them using Type Families, or using Type Classes with Functional Dependencies.
Returning to the first part of your question, Beta Reduction is the process of filling in concrete values for each of the function's arguments. I've heard people describe expressions as "after reduction" or "fully reduced" to emphasize some stage in this process. This is similar to a function Call Site, but emphasizes the expression & arguments, rather than the surrounding context.
Hello I started to write in sml and I have some difficulty in understanding a particular function.
I have this function:
fun isInRow (r:int) ((x,y)) = x=r;
I would be happy to get explain to some points:
What the function accepts and what it returns.
What is the relationship between (r: int) ((x, y)).
Thanks very much !!!
The function isInRow has two arguments. The first is named r. The second is a pair (x, y). The type ascription (r: int) says that r must be an int.
This function is curried, which is a little unusual for SML. What this means roughly speaking is that it accepts arguments given separately rather than supplied as a pair.
So, the function accepts an int and a pair whose first element is an int. These are accepted as separate arguments. It returns a boolean value (the result of the comparison x = r).
A call to the function would look like this:
isInRow 3 (3, 4)
There is more to say about currying (which is kind of cool), but I hope this is enough to get you going.
In addition to what Jeffrey has said,
You don't need the extra set of parentheses:
fun isInRow (r:int) (x,y) = x=r;
You don't need to specify the type :int. If you instead write:
fun isInRow r (x,y) = x=r;
then the function's changes type from int → (int • 'a) → bool into ''a → (''a • 'b) → bool, meaning that r and x can have any type that can be compared for equality (not just int), and y can still be anything since it is still disregarded.
Polymorphic functions are one of the strengths of typed, functional languages like SML.
You could even refrain from giving y a name:
fun isInRow r (x,_) = x=r;
[...] a pair of functions tofun : int -> ('a -> 'a) and fromfun : ('a -> 'a) ->
int such that (fromfun o tofun) n evaluates to n for every n : int.
Anyone able to explain to me what this is actually asking for? I'm looking for more of an explanation of that than an actual solution to this.
What this is asking for is:
1) A higher-order function tofun which when given an integer returns a polymorphic function, one which has type 'a->'a, meaning that it can be applied to values of any type, returning a value of the same type. An example of such a function is:
- fun id x = x;
val id = fn : 'a -> 'a
for example, id "cat" = "cat" and id () = (). The later value is of type unit, which is a type with only 1 value. Note that there is only 1 total function from unit to unit (namely, id or something equivalent). This underscores the difficulty with coming up with defining tofun: it returns a function of type 'a -> 'a, and other than the identity function it is hard to think of other functions. On the other hand -- such functions can fail to terminate or can raise an error and still have type 'a -> 'a.
2) fromfun is supposed to take a function of type 'a ->'a and return an integer. So e.g. fromfun id might evaluate to 0 (or if you want to get tricky it might never terminate or it might raise an error)
3) These are supposed to be inverses of each other so that, e.g. fromfun (tofun 5) needs to evaluate to 5.
Intuitively, this should be impossible in a sufficiently pure functional language. If it is possible in SML, my guess is that it would be by using some of the impure features of SML (which allow for side effects) to violate referential transparency. Or, the trick might involve raising and handling errors (which is also an impure feature of SML). If you find an answer which works in SML it would be interesting to see if it could be translated to the annoyingly pure functional language Haskell. My guess is that it wouldn't translate.
You can devise the following property:
fun prop_inverse f g n = (f o g) n = n
And with definitions for tofun and fromfun,
fun tofun n = ...
fun fromfun f = ...
You can test that they uphold the property:
val prop_test_1 =
List.all
(fn i => prop_inverse fromfun tofun i handle _ => false)
[0, ~1, 1, valOf Int.maxInt, valOf Int.minInt]
And as John suggests, those functions must be impure. I'd also go with exceptions.
Imagine I have a custom type and two functions:
type MyType = Int -> Bool
f1 :: MyType -> Int
f3 :: MyType -> MyType -> MyType
I tried to pattern match as follows:
f1 (f3 a b i) = 1
But it failed with error: Parse error in pattern: f1. What is the proper way to do the above?? Basically, I want to know how many f3 is there (as a and b maybe f3 or some other functions).
You can't pattern match on a function. For (almost) any given function, there are an infinite number of ways to define the same function. And it turns out to be mathematically impossible for a computer to always be able to say whether a given definition expresses the same function as another definition. This also means that Haskell would be unable to reliably tell whether a function matches a pattern; so the language simply doesn't allow it.
A pattern must be either a single variable or a constructor applied to some other patterns. Remembering that constructor start with upper case letters and variables start with lower case letters, your pattern f3 a n i is invalid; the "head" of the pattern f3 is a variable, but it's also applied to a, n, and i. That's the error message you're getting.
Since functions don't have constructors, it follows that the only pattern that can match a function is a single variable; that matches all functions (of the right type to be passed to the pattern, anyway). That's how Haskell enforces the "no pattern matching against functions" rule. Basically, in a higher order function there's no way to tell anything at all about the function you've been given except to apply it to something and see what it does.
The function f1 has type MyType -> Int. This is equivalent to (Int -> Bool) -> Int. So it takes a single function argument of type Int -> Bool. I would expect an equation for f1 to look like:
f1 f = ...
You don't need to "check" whether it's an Int -> Bool function by pattern matching; the type guarantees that it will be.
You can't tell which one it is; but that's generally the whole point of taking a function as an argument (so that the caller can pick any function they like knowing that you'll use them all the same way).
I'm not sure what you mean by "I want to know how many f3 is there". f1 always receives a single function, and f3 is not a function of the right type to be passed to f1 at all (it's a MyType -> MyType -> MyType, not a MyType).
Once a function has been applied its syntactic form is lost. There is now way, should I provide you 2 + 3 to distinguish what you get from just 5. It could have arisen from 2 + 3, or 3 + 2, or the mere constant 5.
If you need to capture syntactic structure then you need to work with syntactic structure.
data Exp = I Int | Plus Exp Exp
justFive :: Exp
justFive = I 5
twoPlusThree :: Exp
twoPlusThree = I 2 `Plus` I 3
threePlusTwo :: Exp
threePlusTwo = I 2 `Plus` I 3
Here the data type Exp captures numeric expressions and we can pattern match upon them:
isTwoPlusThree :: Exp -> Bool
isTwoPlusThree (Plus (I 2) (I 3)) = True
isTwoPlusThree _ = False
But wait, why am I distinguishing between "constructors" which I can pattern match on and.... "other syntax" which I cannot?
Essentially, constructors are inert. The behavior of Plus x y is... to do nothing at all, to merely remain as a box with two slots called "Plus _ _" and plug the two slots with the values represented by x and y.
On the other hand, function application is the furthest thing from inert! When you apply an expression to a function that function (\x -> ...) replaces the xes within its body with the applied value. This dynamic reduction behavior means that there is no way to get a hold of "function applications". They vanish into thing air as soon as you look at them.
cube (x,y,z) =
filter (pcubes x) cubes
cubes = [(a,b,c) | a <- [1..30],b <- [1..30],c <- [1..30]]
pcubes x (b,n,m) = (floor(sqrt(b*n)) == x)
so this code works, cubes makes a list of tuples,pcubes is used with filter to filter all the cubes in which floor(sqrt(b*n)) == x is satisfied,but the person who has modified my code wrote pcubes x in filter (pcubes x) cubes,how does this work.pcubes x makes a function that will initial the cubes x (b,n,m) that will take in a tuple and output a bool.the bool will be used in the filter function. How does this sort of manipulation happen? how does pcubes x access the (b,n,m) part of the function?
In Haskell, we don't usually use tuples (ie: (a,b,c)) to pass arguments to functions. We use currying.
Here's an example:
add a b = a + b
Here add is a function that takes a number, the returns another function that takes a number, then returns a number. We represent it's type as so:
add :: Int -> (Int -> Int)
Because of the way -> behaves, we can remove the parentheses in this case:
add :: Int -> Int -> Int
It is called like this:
(add 1) 2
but because of the way application works, we can just write:
add 1 2
Doesn't that look like our definition above, of the form add a b...?
Your function pcubes is similar. Here's how I'd write it:
pcubes x (b,n,m) = floor (sqrt (b*n)) == x
And as someone else said, it's type could be represented as:
pcubes :: Float -> (Float, Float, Float) -> Bool
When we write pcubes 1 the type becomes:
pcubes 1 :: (Float, Float, Float) -> Bool
Which, through currying, is legal, and can quite happily be used elsewhere.
I know, this is crazy black functional magic, as it was for me, but before long I guarantee you'll never want to go back: curried functions are useful.
A note on tuples: Expressions like (a,b,c) are data . They are not purely function-argument expressions. The fact that we can pull it into a function is called pattern matching, though it's not my turn to go into that.