Welcome to OGeek Q&A Community for programmer and developer-Open, Learning and Share
Welcome To Ask or Share your Answers For Others

Categories

0 votes
577 views
in Technique[技术] by (71.8m points)

recursion - F#: Recursive Functions: concatenate 2 lists which have common elements

So here is what I have so far. It feels close but im not sure how to fix the problems in line 84 (2nd to last line: elif List.append(isolate(a),isolate(b)) != [] then List.append(isolate(a),isolate(b))).

    (* val isolate : l:'a list -> 'a list when 'a : equality *)
    let rec isolate (l:'a list) = 
        match l with
        | [] -> []
        | x::xs ->
            if memberof(x,xs)
            then
                let xs = remove (x,l)
                isolate xs
            else isolate xs
    ( * val common : 'a list * 'a list -> 'a list when 'a : equality *)
    let rec common (k: 'a list, l:'a list) = 
        match ((k:'a list),(l:'a list)) with
        | (a, b) -> 
            if a=[] then []
            elif b=[] then []
            elif List.append(isolate(a),isolate(b)) != [] then List.append(isolate(a),isolate(b))
            else []

edit:

asked to post whole code:

(* val sumlist : l:float list -> float *)
let rec sumlist l =  
    match (l:float list) with
    | [] -> 0.0
    | a::x -> (sumlist x) + a
    (* :: creates a list. *)
sumlist([1.0;2.0;3.0])
(* val squarelist : l:float list -> float list *)
let rec squarelist l = 
    match (l:float list) with
    | [] -> []
    | a::x -> (a*a)::(squarelist x)

(* val mean : l:float list -> float *)
let mean l = 
    match (l:float list) with
    | [] -> 0.0
    | l -> (sumlist l)/(float)l.Length
mean([1.0;2.0;3.0])
(* val mean_diffs : l:float list -> float list *)
let mean_diffs l = 
    match l with
    set a = mean(l)

    | [] -> []
    let rec diffs (a,l)=
        match l with
        | x::xs -> (x-(mean(l))::diffs(xs)
        | [] -> l
mean_diffs([1.0;2.0;3.0])

(* val variance : l:float list -> float *)
let variance l = 
    match (l:float list) with
    | [] -> 0.0
    | l -> (sumlist (squarelist (mean_diffs l)))/(float)l.Length


(* End of question 1 *) (* Do not edit this line. *)

(* Question 2 *) (* Do not edit this line. *)

(* val memberof : 'a * 'a list -> bool when 'a : equality *)
let rec memberof l=
    match (l: 'a * 'a list) with
    | (t,[]) -> false
    | (t, x::xs) when t=x -> true
    | (t, x::xs) -> t=x || memberof(t,xs)

(* val remove : 'a * 'a list -> 'a list when 'a : equality *)
let rec remove ((k:'a),(l:'a list)) = 
    match l with
    | [] -> []
    | x::xs when x=k -> xs
    | x::xs ->x::(remove(k,xs))

(* End of question 2 *) (* Do not edit this line *)

(* Question 3 *) (* Do not edit this line *)

(* val isolate : l:'a list -> 'a list when 'a : equality *)
let rec isolate (l:'a list) = 
    match l with
    | [] -> []
    | x::xs ->
        if memberof(x,xs)
        then
            let xs = remove (x,l)
            isolate xs
        else isolate xs

(* End of question 3 *) (* Do not edit this line *)

(* Question 4 *) (* Do not edit this line *)

(* val common : 'a list * 'a list -> 'a list when 'a : equality *)
let rec common (k: 'a list, l:'a list) = 
    match ((k:'a list),(l:'a list)) with
    | (a, b) -> 
        if a=[] then []
        elif b=[] then []    
        elif List.append(isolate(a),isolate(b)) <> [] then List.append(isolate(a),isolate(b))
        else []
common([1.0;2.0;6.0;10.0],[5.0;6.0;10.0])

It seems that <> has fixed the problem but do you have any advice on my function mean_diffs?

See Question&Answers more detail:os

与恶龙缠斗过久,自身亦成为恶龙;凝视深渊过久,深渊将回以凝视…
Welcome To Ask or Share your Answers For Others

1 Reply

0 votes
by (71.8m points)

Since this appears that you are working on a course and it is building upon the previous exercises, the code is converted to more F# idiomatic and a standardized format of recursive functions to make them easier to use when you get to currying See: F# for fun and profit and Functions as First-Class Values (F#) and other more advanced concepts.

The format is basically

let funXYZ list =
    let rec funXYZInner list acc =
        match list with
        | head :: tail ->
            let acc = (somefunc head) :: acc
            funXYZInner tail acc
        | [] -> acc
    funXYZInner list []

where funXYZ is an exposed function name that does NOT have a rec. I can't recall the source but if you can implement a function needing a rec with the rec not being exposed it makes the code more portable.

The basic concept is that you take a list and pull the list apart into a head and tail:

head :: tail

then you process the head:

somefunc head

then accumulate the result into the accumulator

let acc = value :: acc
let acc = value + acc
let acc = acc + (value * value)

then process the remainder of the list, e.g. tail, passing the accumulator

funXYZInner tail acc

When the input list matches empty

| []

just return the result which was accumulated in the accumulator

acc

The inner function funXYZInner does have a rec and uses an accumulator, i.e. acc. This will help in understanding how to use tail calls which will keep you from running out of memory on large computations.

You probably know already that with match statements you want to cover all of the cases of the match variable. This is because of algebraic data types and is the reason you see those warnings about not all cases being covered. If you see one of those warnings and don't know why you are getting it you need to fix them, or run the risk of unexpected runtime errors or crashes.

While the code you gave can only work with float type list, in the future to make it work with more types you will need to learn about LanguagePrimitives.GenericZero<^T> Type Function (F#).

There are some more basic functions that were added because they were needed, e.g. reverse, and help to show the progression as the examples get more complex.

Do to the fact that examples build upon themselves and you had a specific error in the last one, it was better to give the examples a better foundation which should mitigate common problems encountered when first learning about recursive functions.

With regards to the accumulator, the accumulator can hold different types, e.g. float, list, int, and there can be more than one accumulator being used in the recursive function, e.g. numeratorAcc, denominatorAcc. Also by pulling out the calculation of the accumulator value, e.g.let acc = ..., when you get to more advanced functions you can just pass in a function to replace that calculation.

There is one predicate function memberof which does not use an accumulator. A predicate is a function that returns true or false, and once you reach the desired value you can stop processing the remainder of the list.

Also of note is that while some of the functions could call earlier defined functions, the examples don't make the calls so that they can process the list in one pass. When functions call other functions with list, each function has to process the entire list to return the result. By using rec functions it is sometimes possible to process the list once by doing multiple calculations with the head. However there are times this cannot be done. I did not maximize the functions one way or the other but left them a way that give more variation for learning. Feel free to rewrite them which will lead to function composition.

You will probably have more questions about these examples, so please ask as separate SO questions instead of building on this one.

All the code

// val reverse : l:'a list -> 'a list
let reverse l =
    let rec reverseInner l acc =
        match l with
        | x::xs -> 
            let acc = x :: acc
            reverseInner xs acc
        | [] -> acc
    reverseInner l []

reverse [ 3.0; 2.0; 1.0 ]  // val it : float list = [1.0; 2.0; 3.0]    

// val length : l:'a list -> int
let length l =
    let rec lengthInner l acc =
        match l with
        | x::xs -> 
            let acc = acc + 1
            lengthInner xs acc
        | [] -> acc
    lengthInner l 0

length [ 3.0; 2.0; 1.0 ]  // val it : int = 3

// val sum : l:float list -> float
let sum l =
    let rec sumInner l acc =
        match l with
        | x::xs -> 
            let acc = acc + x
            sumInner xs acc
        | [] -> acc
    sumInner l 0.0

sum [ 1.0; 2.0; 3.0 ]  // val it : float = 6.0

// val square : l:float list -> float list
let square (l : float list) = 
    let rec squareInner l acc =
        match l with
        | x::xs -> 
            let acc = (x * x) :: acc
            squareInner xs acc
        | [] -> reverse acc
    squareInner l []

square [ 1.0; 2.0; 3.0 ]  // val it : float list = [1.0; 4.0; 9.0]

// val mean : l:float list -> float
let mean l = 
    let rec meanInner l sumacc lengthacc =
        match l with
        | x::xs -> 
            let sumacc = sumacc + x
            let lengthacc = lengthacc + 1.0
            meanInner xs sumacc lengthacc
        | [] -> sumacc / lengthacc
    meanInner l 0.0 0.0

mean([1.0;2.0;3.0]) // val it : float = 2.0

// val mean_diffs : l:float list -> float list
let meanDiff l = 
    let rec meanDiffInner l m acc =
        match l with
        | x::xs -> 
            let diff = (x - m)
            let acc = diff :: acc
            meanDiffInner xs m acc
        | [] -> reverse acc
    meanDiffInner l (mean l) []

meanDiff [ 1.0; 2.0; 3.0 ]  // val it : float list = [-1.0; 0.0; 1.0]


// From: https://en.wikipedia.org/wiki/Variance
// Suppose a population of numbers consists of 3, 4, 7, and 10. 
// The arithmetic mean of these numbers, often informally called the "average", is (3+4+7+10)÷4 = 6. 
// The variance of these four numbers is the average squared deviation from this average. 
// These deviations are (3–6) = –3, (4–6) = –2, (7–6) = 1, and (10–6) = 4. 
// Thus the variance of the four numbers is ((-3)^2 + (-2)^2 + (1)^2 + (4)^2) / 4 = 15/2 = 7.5

// val variance : l:float list -> float
let variance l = 
    let deviations = meanDiff l
    let rec varianceInner l numeratorAcc denomenatorAcc =
        match l with
        | devation::xs ->
            let numeratorAcc = numeratorAcc + (devation * devation) 
            let denomenatorAcc = denomenatorAcc + 1.0
            varianceInner xs numeratorAcc denomenatorAcc
        | [] -> numeratorAcc / denomenatorAcc
    varianceInner deviations 0.0 0.0

variance [ 1.0; 2.0; 3.0 ]          // val it : float = 0.6666666667
variance [ 3.0; 4.0; 7.0; 10.0 ]    // val it : float = 7.5


(* End of question 1 *) (* Do not edit this line. *)

(* Question 2 *) (* Do not edit this line. *)

// val memberof : l:'a list -> item:'a -> bool when 'a : equality
let memberof l item =
    let rec memberInner l item =
        match l with
        | x::xs -> 
            if x = item then
                true
            else 
                memberInner xs item
        | [] -> false
    memberInner l item

memberof [ 1.0; 2.0; 3.0 ] 0.0  // val it : bool = false
memberof [ 1.0; 2.0; 3.0 ] 1.0  // val it : bool = true
memberof [ 1.0; 2.0; 3.0 ] 2.0  // trueval it : bool = true
memberof [ 1.0; 2.0; 3.0 ] 3.0  // val it : bool = true
memberof [ 1.0; 2.0; 3.0 ] 4.0  // val it : bool = false

// val remove : l:'a list -> item:'a -> 'a list when 'a : equality
let remove l item =
    let rec removeInner l item acc =
        match l with
        | x::xs ->
            if x = item then
                removeInner xs item acc
            else
                let acc = x :: acc
                removeInner xs item acc
        | [] -> reverse acc
    removeInner l item []

remove [ 1.0; 2.0; 3.0 ] 0.0    // val it : float list = [1.0; 2.0; 3.0]
remove [ 1.0; 2.0; 3.0 ] 1.0    // val it : float list = [2.0; 3.0]
remove [ 1.0; 2.0; 3.0 ] 2.0    // val it : float list = [1.0; 3.0]
remove [ 1.0; 2.0; 3.0 ] 3.0    // val it : float list = [1.0; 2.0]
remove [ 1.0; 2.0; 3.0 ] 4.0    // val it : float list = [1.0; 2.0; 3.0]

(* End of question 2 *) (* Do not edit this line *)

(* Question 3 *) (* Do not edit this line *)

// val isolate : list:'a list -> 'a list when 'a : equality
let isolate list =
    let rec isolateInner searchList commonlist =
        match searchList with
        | x::xs ->
            if (memberof commonlist x) then
                isolateInner xs commonlist
            else
                let commonlist = (x :: commonlist)
                isolateInner xs commonlist
        | [] -> reverse commonlist
    isolateInner list []

isolate [ 1.0; 2.0; 3.0 ]                                    // val it : float list = [1.0; 2.0; 3.0]
isolate [ 1.0; 1.0; 2.0; 3.0 ]                               // val it : float list = [1.0; 2.0; 3.0]
isolate [ 1.0; 2.0; 2.0; 3.0 ]                               // val it : float list = [1.0; 2.0; 3.0]
isolate [ 1.0; 2.0; 3.0; 3.0 ]                               // val it : float list = [1.0; 2.0; 3.0]
isolate [ 3.0; 2.0; 1.0; 1.0; 2.0; 3.0; 2.0; 1.0; 1.0; 3.0]  // val it : float list = [3.0; 2.0; 1.0]

(* End of question 3 *) (* Do not edit this line *)

(* Question 4 *) (* Do not edit this line *)

// val common : a:'a list -> b:'a list -> 'a list when 'a : equality
let common a b = 
    let rec commonInner a b acc =
        match (a,b) with
        | (x::xs,b) ->
            if (memberof acc x) then
                commonInner xs b acc
            else
                let acc = x :: acc
                commonInner xs b acc
        | ([],y::ys) ->
            if (memberof acc y) then
                commonInner [] ys acc
            else
                let acc = y :: acc
                commonInner [] ys acc
        | ([],[]) -> reverse acc
    commonInner a b []

common [ 1.0; 2.0; 6.0; 10.0] [ 5.0; 6.0; 10.0 ]   // val it : float list = [1.0; 2.0; 6.0; 10.0; 5.0]

与恶龙缠斗过久,自身亦成为恶龙;凝视深渊过久,深渊将回以凝视…
OGeek|极客中国-欢迎来到极客的世界,一个免费开放的程序员编程交流平台!开放,进步,分享!让技术改变生活,让极客改变未来! Welcome to OGeek Q&A Community for programmer and developer-Open, Learning and Share
Click Here to Ask a Question

...