7 questions to apply what we've seen regarding functions
1. How to define the return value (v) of a function (f)? ⏱ 10''
A. Simply name the value result.
B. End the function with return v.
C. Do f = v
D.v is the last line of f.
Answer
A. ❌ F♯ is not a convention-based language.
B. ❌ F♯ does have a return keyword, but it's used in computation expressions. 📍
C. ❌ This is the VB way, not the F♯ way.
D. ✅ The returned value of a function is its whole body: an expression that can be composed from sub-expressions.
2. How to write an add function taking 2 strings and returning an int? ⏱ 20’’
A.let add a b = a + b
B.let add (a: string) (b: string) = (int a) + (int b)
C.let add (a: string) (b: string) : int = a + b
Answer
A. ❌
+ operator works on different types.
There is no way to define a type for all the types supported by +.
The compiler only takes the first type, which is int.
In let add a b = a + b, a and b are inferred to be of type int.
B. ✅
The type of a and b must be specified.
They must be converted to int.
The int return type can be inferred.
C. ❌
Returns a string
3. What does this code: add >> multiply? ⏱ 10’’
A. Create a pipeline
B. Define a named function
C. Compose 2 functions
Answer
A. ❌ Pipelines are based on the pipe operator |>
B. ❌ A named function is defined with a let binding. add >> multiply is an anonymous function.
C. ✅ >> is the compose operator, taking 2 functions as arguments to build a new function.
A.let ? _ = ()
B.let ? x = x
C.let ? f x = f x
D.let ? x f = f x
E.let ? f g x = g (f x)
💡 Tips: These may be operators.
Answer
A.let ? _ = ()
B.let ? x = x
C.let ? f x = f x
This function is taking 2 parameters f and x.
f x implies that f is a function.
This function as a function adds no value.
On the contrary, as operator, we get the benefit of having the possibility to use the operator between the function and its value, for instance to remove the need to use parentheses around the expression used to be passed as argument to the function.
This operator is called pipe left.
D.let ? x f = f x
Closed to the previous function, with the 2 parameters swapped.
Similarly, it makes sense only as an operator, called pipe right or just pipe.
E.let ? f g x = g (f x)
This function is taking 3 parameters f, g and x.
f x and g (...) indicates that f and g are functions.
We call f with the argument x, then we call g with the previous result.
It's the definition of the compose right operator, noted >>.
5. Describe the following functions in terms of the number of parameters, their type, the return type ⏱ 60’’
A.int -> unit
B.unit -> int
C.string -> string -> string
D.('T -> bool) -> 'T list -> 'T list
Answer
A.int -> unit
1 parameter: int
no return value
B.unit -> int
no parameter
returns an int
C.string -> string -> string
2 parameters, both string
returns a string
→ Can be the + operator.
D.('T -> bool) -> 'T list -> 'T list
2 parameters
('T -> bool) is a function returning a boolean
→ It's a predicate.
'T list is a list.
returns a list
→ Can be the List.filter function.
6. What is the signature of the h function below? ⏱ 30’’
let f x = x + 1
let g x y = $"%i{x} + %i{y}"
let h = f >> g
A.int -> int
B.int -> string
C.int -> int -> string
D.int -> int -> int
Answer
C.int -> int -> string ✅
+ 1 → x: int
→ f: (x: int) -> int
%i{x} and %i{y} in interpolated string $"..." → x and y are ints
→ (x: int) -> (y: int) -> string
>> compose operator
→ h takes the same parameter as f and returns what is returned by the call to g with the value returned from f
→ But, g takes 2 parameters. Hence, g is partially applied and 1 argument is still needed.
The exact signature is in fact int -> (int -> string).
☝️ Notes:
This question was hard enough to illustrate the misuse of >> because composed functions have different arities (f has 1 parameter, g has 2).
It's way simpler to write let h x y = g (f x) y.
7. What value returns f 2? ⏱ 10’’
let f = (-) 1
f 2;; // ?
A.1
B.3
C.-1
Answer
C.-1 ❗
Counter-intuitive, isn't it? We expect f to decrement by 1.
It's easier to understand what's going on if we write f like this: let f x = 1 - x.
💡 Hint: the function that decrements by 1 can be written as:
let f = (+) -1(this works here because + is commutative whereas - is not)
