Programming Puzzles & Code Golf Stack Exchange is a question and answer site for programming puzzle enthusiasts and code golfers. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

This isn't very widely known, but what we call the Fibonacci sequence, AKA

1, 1, 2, 3, 5, 8, 13, 21, 34...

is actually called the Duonacci sequence. This is because to get the next number, you sum the previous 2 numbers. There is also the Tribonacci sequence,

1, 1, 1, 3, 5, 9, 17, 31, 57, 105, 193, 355, 653, 1201...

because the next number is the sum of the previous 3 numbers. And the Quadronacci sequence

1, 1, 1, 1, 4, 7, 13, 25, 49, 94, 181, 349, 673...

And everybody's favorite, the Pentanacci sequence:

1, 1, 1, 1, 1, 5, 9, 17, 33, 65, 129...

And the Hexanacci sequence, the Septanacci sequence, the Octonacci sequence, and so on and so forth up to the N-Bonacci sequence.

The Challenge

You must write a function or program that takes two numbers N and X, and prints out the first X N-Bonacci numbers. N will be a whole number larger than 0, and you can safely assume no N-Bonacci numbers will exceed the default number type in your language. The output can be in any human readable format, and you can take input in any reasonable manner. (Command line arguments, function arguments, STDIN, etc.)

As usual, this is Code-golf, so standard loopholes apply and the shortest answer in bytes wins!

Sample IO

#n,  x,     output
 3,  8  --> 1, 1, 1, 3, 5, 9, 17, 31
 7,  13 --> 1, 1, 1, 1, 1, 1, 1, 7, 13, 25, 49, 97, 193
 1,  20 --> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1
 30, 4  --> 1, 1, 1, 1       //Since the first 30 are all 1's
 5,  11 --> 1, 1, 1, 1, 1, 5, 9, 17, 33, 65, 129
share|improve this question
    
Man, I had this idea a while ago and never got around to writing it up. – Morgan Thrapp Jan 29 at 21:43
8  
My vote button == your avatar – ETHproductions Jan 29 at 23:50
    
@ETHproductions Haha, yeah that was the inspiration. I also changed my username to fit better with the "goat" theme. – Dr Green Eggs and Ham DJ Jan 29 at 23:51
    
So you used to be DJ McMayhem? I get confused with username changes... – Eᴀsᴛᴇʀʟʏ Iʀᴋ Jan 30 at 1:23
1  
Bonus challenge: given a list of numbers, figure out which "nacci" sequence it is a subset of – Cyoce Jan 30 at 3:08

18 Answers 18

up vote 12 down vote accepted

Boolfuck, 6 bytes

,,[;+]

You can safely assume no N-Bonacci numbers will exceed the default number type in your language.

The default number type in Boolfuck is a bit. Assuming this also extends to the input numbers N and X, and given that N>0, there are only two possible inputs - 10 (which outputs nothing) and 11 (which outputs 1).

, reads a bit into the current memory location. N is ignored as it must be 1. If X is 0, the loop body (surrounded by []) is skipped. If X is 1, it is output and then flipped to 0 so the loop does not repeat.

share|improve this answer

Pyth, 13

<Qu+Gs>QGEm1Q

Test Suite

Takes input newline separated, with n first.

Explanation:

<Qu+Gs>QGEm1Q  ##  implicit: Q = eval(input)
  u      Em1Q  ##  read a line of input, and reduce that many times starting with
               ##  Q 1s in a list, with a lambda G,H
               ##  where G is the old value and H is the new one
   +G          ##  append to the old value
     s>QG      ##  the sum of the last Q values of the old value
<Q             ##  discard the last Q values of this list
share|improve this answer
1  
Wow, that was fast. I barely had time to close my browser before you'd already posted this! – Dr Green Eggs and Ham DJ Jan 29 at 21:51

Python 2, 79 bytes

n,x=input()
i,f=0,[]
while i<x:v=[sum(f[i-n:]),1][i<n];f.append(v);print v;i+=1

Try it online

share|improve this answer
4  
Welcome to PPCG! – VTCAKAVSMoACE Jan 29 at 22:31
    
Try replacing the last line with exec"v=[sum(f[i-n:]),1][i<n];f+=[v];print v;i+=1;"*x – Cyoce Jan 30 at 5:33

Haskell, 56 bytes

g l=sum l:g(sum l:init l)
n#x|i<-1<$[1..n]=take x$i++g i

Usage example: 3 # 8-> [1,1,1,3,5,9,17,31].

How it works

i<-1<$[1..n]           -- bind i to n copies of 1
take x                 -- take the first x elements of
       i++g i          -- the list starting with i followed by (g i), which is
sum l:                 -- the sum of it's argument followed by
      g(sum l:init l)  -- a recursive call to itself with the the first element
                       -- of the argument list replaced by the sum
share|improve this answer
    
Shouldn't that be tail l instead of init l? – proud haskeller Jan 31 at 20:49
    
@proudhaskeller: it doesn't matter. we keep the last n elements int the list. There's no difference between removing from the end and adding to the front and the other way around, i.e. removing from the front and adding to the end, because the initial list is made up of only 1s. – nimi Jan 31 at 21:19
    
Oh, I get it. That's a nifty way to replace ++[] by :! – proud haskeller Jan 31 at 21:26
    
@proudhaskeller: yes, exactly! – nimi Jan 31 at 21:30

Jelly, 12 bytes

ḣ³S;
b1Ç⁴¡Uḣ

Try it online!

How it works

b1Ç⁴¡Uḣ  Main link. Left input: n. Right input: x.

b1       Convert n to base 1.
    ¡    Call...
  Ç        the helper link...
   ⁴       x times.
     U   Reverse the resulting array.
      ḣ  Take its first x elements.


ḣ³S;     Helper link. Argument: A (list)

ḣ³       Take the first n elements of A.
  S      Compute their sum.
   ;     Prepend the sum to A.
share|improve this answer

Python 2, 55 bytes

def f(x,n):l=[1]*n;exec"print l[0];l=l[1:]+[sum(l)];"*x

Tracks a length-n window of the sequence in the list l, updated by appending the sum and removing the first element. Prints the first element each iteration for x iterations.

A different approach of storing all the elements and summing the last n values gave the same length (55).

def f(x,n):l=[1]*n;exec"l+=sum(l[-n:]),;"*x;print l[:x]
share|improve this answer

ES6, 66 bytes

(i,n)=>[...Array(n)].map((_,j,a)=>a[j]=j<i?1:j-i?s+=s-a[j+~i]:s=i)

Sadly map won't let you access the result array in the callback.

share|improve this answer

Javascript ES6/ES2015, 107 97 85 80 Bytes

Thanks to @user81655, @Neil and @ETHproductions for save some bytes


(i,n)=>eval("for(l=Array(i).fill(1);n-->i;)l.push(eval(l.slice(-i).join`+`));l")

try it online


Test cases:

console.log(f(3,  8))// 1, 1, 1, 3, 5, 9, 17, 31
console.log(f(7,  13))// 1, 1, 1, 1, 1, 1, 1, 7, 13, 25, 49, 97, 193
console.log(f(5,  11))// 1, 1, 1, 1, 1, 5, 9, 17, 33, 65, 129
share|improve this answer
1  
Nice. A couple of golfing tips: for is always better than while, x.split('') -> [...x], ~~a -> +a, n-=1 -> n--, if you enclose the entire function body in an eval you don't need to write return. Also, even shorter than [...'1'.repeat(i)] is Array(i).fill(1) and you can remove the ~~ from a and b. And you're allowed to remove the f=. – user81655 Jan 29 at 23:51
2  
This is what it looks like with my tips (85 bytes): (i,n)=>eval("for(l=Array(i).fill(1);n-->i;)l.push(l.slice(-i).reduce((a,b)=>a+b‌​));l"). I changed the order of statements, combined the n-- into n-i and removed l from the arguments to save a few extra bytes. – user81655 Jan 30 at 0:13
1  
@user81655 I don't get the eval saving; (i,n)=>{for(l=Array(i).fill(1);n-->i;)l.push(l.slice(-i).reduce((a,b)=>a+b));re‌​turn l} is still 85 bytes. – Neil Jan 30 at 0:17
    
@Neil Looks like 86 bytes to me... – user81655 Jan 30 at 0:21
3  
l.slice(-i).reduce((a,b)=>a+b) => eval(l.slice(-i).join`+`) – ETHproductions Jan 30 at 0:52

C++11, 360 bytes

Hi I just like this question. I know c++ is a very hard language to win this competition. But I'll thrown a dime any way.

#include<vector>
#include<numeric>
#include<iostream>
using namespace std;typedef vector<int>v;void p(v& i) {for(auto&v:i)cout<<v<<" ";cout<<endl;}v b(int s,int n){v r(n<s?n:s,1);r.reserve(n);for(auto i=r.begin();r.size()<n;i++){r.push_back(accumulate(i,i+s,0));}return r;}int main(int c, char** a){if(c<3)return 1;v s=b(atoi(a[1]),atoi(a[2]));p(s);return 0;}

I'll leave this as the readable explanation of the code above.

#include <vector>
#include <numeric>
#include <iostream>

using namespace std;
typedef vector<int> vi;

void p(const vi& in) {
    for (auto& v : in )
        cout << v << " ";
    cout << endl;
}

vi bonacci(int se, int n) {
    vi s(n < se? n : se, 1);
    s.reserve(n);
    for (auto it = s.begin(); s.size() < n; it++){
        s.push_back(accumulate(it, it + se, 0));
    }
    return s;
}

int main (int c, char** v) {
    if (c < 3) return 1;
    vi s = bonacci(atoi(v[1]), atoi(v[2]));
    p(s);
    return 0;
}
share|improve this answer
    
Welcome to Programming Puzzles and Code Golf. This is a good answer, however I have noticed that you have lots of whitespace, and variable and function names that are longer than 1 character long. As it stands, this is a good readable version of your code, but you should add a golfed version. When you do, I will give you an upvote, but until it is golfed I will not. – wizzwizz4 Jan 30 at 13:28
    
@wizzwizz4 Hi, added a golfed version of the code above. I left the ungolfed code around to let people see how I did it. Besides I like to read a function bonacci that returns vi which still sounds like vibonacci. I do feel I should not make the main function shorter because the standardard mandates using int main(int, char**) as entry point of the program. Further I believe all variables are max 1 character long and all non significant whitespaces are removed. – hetepeperfan Jan 30 at 14:36
1  
This is not code-"comply with the standards". This is code-golf. We manipulate and take advantage of our languages. If any variables are ints, remove the int. If any functions are called foo, call them f. Be brutal; ignore the standard and exploit the compiler. That is how you golf. – wizzwizz4 Jan 30 at 16:11
    
Puns and nice code belong in the ungolfed code only. But feel free to keep them there; actually, it is recommended to. But be really, really mean to the compiler when you golf your code. Get it as small as possible no matter what. (Oh, and here's the +1 I promised!) – wizzwizz4 Jan 30 at 16:12
    
@wizzwizz4 Is removing "int" valid? I thought implied int won't run. – Dr Green Eggs and Ham DJ Jan 30 at 20:18

Python 3, 59

Saved 20 bytes thanks to FryAmTheEggman.

Not a great solution, but it'll work for now.

def r(n,x):f=[1]*n;exec('f+=[sum(f[-n:])];'*x);return f[:x]

Also, here are test cases:

assert r(3, 8) == [1, 1, 1, 3, 5, 9, 17, 31]
assert r(7, 13) == [1, 1, 1, 1, 1, 1, 1, 7, 13, 25, 49, 97, 193]
assert r(30, 4) == [1, 1, 1, 1]
share|improve this answer

APL, 21

{⍵↑⍺{⍵,+/⍺↑⌽⍵}⍣⍵+⍺/1}

This is a function that takes n as its left argument and x as its right argument.

Explanation:

{⍵↑⍺{⍵,+/⍺↑⌽⍵}⍣⍵+⍺/1}
                   ⍺/1  ⍝ begin state: X ones    
                  +     ⍝ identity function (to separate it from the ⍵)
    ⍺{         }⍣⍵     ⍝ apply this function N times to it with X as left argument
      ⍵,               ⍝ result of the previous iteration, followed by...
        +/              ⍝ the sum of
          ⍺↑            ⍝ the first X of
            ⌽          ⍝ the reverse of
             ⍵         ⍝ the previous iteration
 ⍵↑                    ⍝ take the first X numbers of the result

Test cases:

      ↑⍕¨ {⍵↑⍺{⍵,+/⍺↑⌽⍵}⍣⍵+⍺/1} /¨ (3 8)(7 13)(1 20)(30 4)(5 11)
 1 1 1 3 5 9 17 31                       
 1 1 1 1 1 1 1 7 13 25 49 97 193         
 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 
 1 1 1 1                                 
 1 1 1 1 1 5 9 17 33 65 129              
share|improve this answer

Julia, 78 bytes

f(n,x)=(z=ones(Int,n);while endof(z)<x push!(z,sum(z[end-n+1:end]))end;z[1:x])

This is a function that accepts two integers and returns an integer array. The approach is simple: Generate an array of ones of length n, then grow the array by adding the sum of the previous n elements until the array has length x.

Ungolfed:

function f(n, x)
    z = ones(Int, n)
    while endof(z) < x
        push!(z, sum(z[end-n+1:end]))
    end
    return z[1:x]
end
share|improve this answer

Haskell, 54 bytes

n%i|i<=n=1|0<1=sum$map(n%)[i-n..i-1]
n?x=map(n%)[1..x]

Defines the binary function ?, used like 3?8 == [1,1,1,3,5,9,17,31].

The auxiliary function ? recursively finds the ith element of the n-bonacci sequence by summing the previous n values. Then, the function ? tabulates the first x values of %.

share|improve this answer

MATL, 22 26 bytes

1tiXIX"i:XK"tPI:)sh]K)

This uses current release (10.2.1) of the language/compiler.

Try it online!

A few extra bytes :-( due to a bug in the G function (paste input; now corrected for next release)

Explanation

1tiXIX"      % input N. Copy to clipboard I. Build row array of N ones
i:XK         % input X. Build row array [1,2,...X]. Copy to clipboard I
"            % for loop: repeat X times. Consumes array [1,2,...X]
  t          % duplicate (initially array of N ones)
  PI:)       % flip array and take first N elements
  sh         % compute sum and append to array
]            % end
K)           % take the first X elements of array. Implicitly display
share|improve this answer

C, 132 bytes

The recursive approach is shorter by a couple of bytes.

k,n;f(i,s,j){for(j=s=0;j<i&j++<n;)s+=f(i-j);return i<n?1:s;}main(_,v)int**v;{for(n=atoi(v[1]);k++<atoi(v[2]);)printf("%d ",f(k-1));}

Ungolfed

k,n; /* loop index, n */

f(i,s,j) /* recursive function */
{
    for(j=s=0;j<i && j++<n;) /* sum previous n n-bonacci values */
        s+=f(i-j);
    return i<n?1:s; /* return either sum or n, depending on what index we're at */
}

main(_,v) int **v;
{
    for(n=atoi(v[1]);k++<atoi(v[2]);) /* print out n-bonacci numbers */
        printf("%d ", f(k-1));
}
share|improve this answer

Java, 82 + 58 = 140 bytes

Function to find the ith n-bonacci number (82 bytes):

int f(int i,int n){if(i<=n)return 1;int s=0,q=0;while(q++<n)s+=f(i-q,n);return s;}

Function to print first k n-bonacci number (58 bytes):

(k,n)->{for(int i=0;i<k;i++){System.out.println(f(i,n));}}
share|improve this answer

Perl 6, 52~72 47~67 bytes

sub a($n,$x){EVAL("1,"x$n~"+*"x$n~"...*")[^$x]}

Requires the module MONKEY-SEE-NO-EVAL, because of the following error:

===SORRY!=== Error while compiling -e
EVAL is a very dangerous function!!! (use MONKEY-SEE-NO-EVAL to override,
but only if you're VERY sure your data contains no injection attacks)
at -e:1

$ perl6 -MMONKEY-SEE-NO-EVAL -e'a(3,8).say;sub a($n,$x){EVAL("1,"x$n~"+*"x$n~"...*")[^$x]}'
(1 1 1 3 5 9 17 31)
share|improve this answer
    
Anyone know of a way to turn off strict mode, etc? – andlrc Jan 29 at 23:34
    
I think if you use a pre-christmas 2015 Perl 6 release, it doesn't enforce monkey-see-no-eval. – Batman Jan 30 at 17:29

Perl 6, 38 bytes

->\N,\X{({@_[*-N..*].sum||1}...*)[^X]} # 38 bytes
-> \N, \X {
  (

    {

      @_[
        *-N .. * # previous N values
      ].sum      # added together

      ||     # if that produces 0 or an error
      1      # return 1

    } ... *  # produce an infinite list of such values

  )[^X]      # return the first X values produced
}

Usage:

# give it a lexical name
my &n-bonacci = >\N,\X{…}

for ( (3,8), (7,13), (1,20), (30,4), (5,11), ) {
  say n-bonacci |@_
}
(1 1 1 3 5 9 17 31)
(1 1 1 1 1 1 1 7 13 25 49 97 193)
(1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1)
(1 1 1 1)
(1 1 1 1 1 5 9 17 33 65 129)
share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.