Zeckendorf number representation
You are encouraged to solve this task according to the task description, using any language you may know.
Just as numbers can be represented in a positional notation as sums of multiples of the powers of ten (decimal) or two (binary); all the positive integers can be represented as the sum of one or zero times the distinct members of the Fibonacci series.
Recall that the first six distinct Fibonacci numbers are: 1, 2, 3, 5, 8, 13.
The decimal number eleven can be written as 0*13 + 1*8 + 0*5 + 1*3 + 0*2 + 0*1 or 010100 in positional notation where the columns represent multiplication by a particular member of the sequence. Leading zeroes are dropped so that 11 decimal becomes 10100.
10100 is not the only way to make 11 from the Fibonacci numbers however; 0*13 + 1*8 + 0*5 + 0*3 + 1*2 + 1*1 or 010011 would also represent decimal 11. For a true Zeckendorf number there is the added restriction that no two consecutive Fibonacci numbers can be used which leads to the former unique solution.
- Task
Generate and show here a table of the Zeckendorf number representations of the decimal numbers zero to twenty, in order.
The intention in this task to find the Zeckendorf form of an arbitrary integer. The Zeckendorf form can be iterated by some bit twiddling rather than calculating each value separately but leave that to another separate task.
- Also see
- OEIS A014417 for the the sequence of required results.
- Brown's Criterion - Numberphile
- Related task
360 Assembly
<lang 360asm>* Zeckendorf number representation 04/04/2017 ZECKEN CSECT
USING ZECKEN,R13 base register B 72(R15) skip savearea DC 17F'0' savearea STM R14,R12,12(R13) save previous context ST R13,4(R15) link backward ST R15,8(R13) link forward LR R13,R15 set addressability LA R6,0 i=0 DO WHILE=(C,R6,LE,=A(20)) do i=0 to 20 MVC PG,=CL80'xx : ' init buffer LA R10,PG pgi=0 XDECO R6,XDEC i MVC 0(2,R10),XDEC+10 output i LA R10,5(R10) pgi+=5 MVC FIB,=A(1) fib(1)=1 MVC FIB+4,=A(2) fib(2)=2 LA R7,2 j=2 LR R1,R7 j SLA R1,2 @fib(j) DO WHILE=(C,R6,GT,FIB-4(R1) do while fib(j)<i LA R7,1(R7) j++ LR R1,R7 j SLA R1,2 ~ L R2,FIB-8(R1) fib(j-1) A R2,FIB-12(R1) fib(j-2) ST R2,FIB-4(R1) fib(j)=fib(j-1)+fib(j-2) LR R1,R7 j SLA R1,2 @fib(j) ENDDO , enddo j LR R8,R6 k=i MVI BB,X'00' bb=false DO WHILE=(C,R7,GE,=A(1)) do j=j to 1 by -1 LR R1,R7 j SLA R1,2 ~ IF C,R8,GE,FIB-4(R1) THEN if fib(j)<=k then MVI BB,X'01' bb=true MVC 0(1,R10),=C'1' output '1' LA R10,1(R10) pgi+=1 LR R1,R7 j SLA R1,2 ~ S R8,FIB-4(R1) k=k-fib(j) ELSE , else IF CLI,BB,EQ,X'01' THEN if bb then MVC 0(1,R10),=C'0' output '0' LA R10,1(R10) pgi+=1 ENDIF , endif ENDIF , endif BCTR R7,0 j-- ENDDO , enddo j IF CLI,BB,NE,X'01' THEN if not bb then MVC 0(1,R10),=C'0' output '0' ENDIF , endif XPRNT PG,L'PG print buffer LA R6,1(R6) i++ ENDDO , enddo i L R13,4(0,R13) restore previous savearea pointer LM R14,R12,12(R13) restore previous context XR R15,R15 rc=0 BR R14 exit
FIB DS 32F Fibonnacci table BB DS X flag PG DS CL80 buffer XDEC DS CL12 temp
YREGS END ZECKEN</lang>
- Output:
0 : 0 1 : 1 2 : 10 3 : 100 4 : 101 5 : 1000 6 : 1001 7 : 1010 8 : 10000 9 : 10001 10 : 10010 11 : 10100 12 : 10101 13 : 100000 14 : 100001 15 : 100010 16 : 100100 17 : 100101 18 : 101000 19 : 101001 20 : 101010
Ada
<lang Ada>with Ada.Text_IO, Ada.Strings.Unbounded;
procedure Print_Zeck is
function Zeck_Increment(Z: String) return String is begin if Z="" then
return "1";
elsif Z(Z'Last) = '1' then
return Zeck_Increment(Z(Z'First .. Z'Last-1)) & '0';
elsif Z(Z'Last-1) = '0' then
return Z(Z'First .. Z'Last-1) & '1';
else -- Z has at least two digits and ends with "10"
return Zeck_Increment(Z(Z'First .. Z'Last-2)) & "00";
end if; end Zeck_Increment; use Ada.Strings.Unbounded; Current: Unbounded_String := Null_Unbounded_String;
begin
for I in 1 .. 20 loop Current := To_Unbounded_String(Zeck_Increment(To_String(Current))); Ada.Text_IO.Put(To_String(Current) & " "); end loop;
end Print_Zeck;</lang>
- Output:
1 10 100 101 1000 1001 1010 10000 10001 10010 10100 10101 100000 100001 100010 100100 100101 101000 101001 101010
ALGOL 68
<lang algol68># print some Zeckendorf number representations #
- We handle 32-bit numbers, the maximum fibonacci number that can fit in a #
- 32 bit number is F(45) #
- build a table of 32-bit fibonacci numbers #
[ 45 ]INT fibonacci; fibonacci[ 1 ] := 1; fibonacci[ 2 ] := 2; FOR i FROM 3 TO UPB fibonacci DO fibonacci[ i ] := fibonacci[ i - 1 ] + fibonacci[ i - 2 ] OD;
- returns the Zeckendorf representation of n or "?" if one cannot be found #
PROC to zeckendorf = ( INT n )STRING:
IF n = 0 THEN "0" ELSE STRING result := ""; INT f pos := UPB fibonacci; INT rest := ABS n; # find the first non-zero Zeckendorf digit # WHILE f pos > LWB fibonacci AND rest < fibonacci[ f pos ] DO f pos -:= 1 OD; # if we found a digit, build the representation # IF f pos >= LWB fibonacci THEN # have a digit # BOOL skip digit := FALSE; WHILE f pos >= LWB fibonacci DO IF rest <= 0 THEN result +:= "0" ELIF skip digit THEN # we used the previous digit # skip digit := FALSE; result +:= "0" ELIF rest < fibonacci[ f pos ] THEN # can't use the digit at f pos # skip digit := FALSE; result +:= "0" ELSE # can use this digit # skip digit := TRUE; result +:= "1"; rest -:= fibonacci[ f pos ] FI; f pos -:= 1 OD FI; IF rest = 0 THEN # found a representation # result ELSE # can't find a representation # "?" FI FI; # to zeckendorf #
FOR i FROM 0 TO 20 DO
print( ( whole( i, -3 ), " ", to zeckendorf( i ), newline ) )
OD </lang>
- Output:
0 0 1 1 2 10 3 100 4 101 5 1000 6 1001 7 1010 8 10000 9 10001 10 10010 11 10100 12 10101 13 100000 14 100001 15 100010 16 100100 17 100101 18 101000 19 101001 20 101010
AppleScript
(mapAccumuL example)
<lang AppleScript>-- ZECKENDORF NUMBERS --------------------------------------------------------
-- zeckendorf :: Int -> String on zeckendorf(n)
script f on |λ|(n, x) if n < x then [n, 0] else [n - x, 1] end if end |λ| end script if n = 0 then {0} as string else item 2 of mapAccumL(f, n, |reverse|(just of tailMay(fibUntil(n)))) as string end if
end zeckendorf
-- fibUntil :: Int -> [Int] on fibUntil(n)
set xs to {} set limit to n script atLimit property ceiling : limit on |λ|(x) (item 2 of x) > (atLimit's ceiling) end |λ| end script script nextPair property series : xs on |λ|([a, b]) set nextPair's series to nextPair's series & b [b, a + b] end |λ| end script |until|(atLimit, nextPair, {0, 1}) return nextPair's series
end fibUntil
-- TEST ---------------------------------------------------------------------- on run
intercalate(linefeed, ¬ map(zeckendorf, enumFromTo(0, 20)))
end run
-- GENERIC FUNCTIONS ---------------------------------------------------------
-- enumFromTo :: Int -> Int -> [Int] on enumFromTo(m, n)
if m > n then set d to -1 else set d to 1 end if set lst to {} repeat with i from m to n by d set end of lst to i end repeat return lst
end enumFromTo
-- foldl :: (a -> b -> a) -> a -> [b] -> a on foldl(f, startValue, xs)
tell mReturn(f) set v to startValue set lng to length of xs repeat with i from 1 to lng set v to |λ|(v, item i of xs, i, xs) end repeat return v end tell
end foldl
-- 'The mapAccumL function behaves like a combination of map and foldl; -- it applies a function to each element of a list, passing an -- accumulating parameter from left to right, and returning a final -- value of this accumulator together with the new list.' (see Hoogle)
-- mapAccumL :: (acc -> x -> (acc, y)) -> acc -> [x] -> (acc, [y]) on mapAccumL(f, acc, xs)
script on |λ|(a, x) tell mReturn(f) to set pair to |λ|(item 1 of a, x) [item 1 of pair, (item 2 of a) & item 2 of pair] end |λ| end script foldl(result, [acc, []], xs)
end mapAccumL
-- map :: (a -> b) -> [a] -> [b] on map(f, xs)
tell mReturn(f) set lng to length of xs set lst to {} repeat with i from 1 to lng set end of lst to |λ|(item i of xs, i, xs) end repeat return lst end tell
end map
-- Lift 2nd class handler function into 1st class script wrapper -- mReturn :: Handler -> Script on mReturn(f)
if class of f is script then f else script property |λ| : f end script end if
end mReturn
-- intercalate :: Text -> [Text] -> Text on intercalate(strText, lstText)
set {dlm, my text item delimiters} to {my text item delimiters, strText} set strJoined to lstText as text set my text item delimiters to dlm return strJoined
end intercalate
-- reverse :: [a] -> [a] on |reverse|(xs)
if class of xs is text then (reverse of characters of xs) as text else reverse of xs end if
end |reverse|
-- tailMay :: [a] -> Maybe [a] on tailMay(xs)
if length of xs > 1 then {nothing:false, just:items 2 thru -1 of xs} else {nothing:true} end if
end tailMay
-- until :: (a -> Bool) -> (a -> a) -> a -> a on |until|(p, f, x)
set mp to mReturn(p) set v to x tell mReturn(f) repeat until mp's |λ|(v) set v to |λ|(v) end repeat end tell return v
end |until|</lang>
- Output:
0 1 10 100 101 1000 1001 1010 10000 10001 10010 10100 10101 100000 100001 100010 100100 100101 101000 101001 101010
AutoHotkey
<lang AutoHotkey>Fib := NStepSequence(1, 2, 2, 20) Loop, 21 { i := A_Index - 1 , Out .= i ":`t", n := "" Loop, % Fib.MaxIndex() { x := Fib.MaxIndex() + 1 - A_Index if (Fib[x] <= i) n .= 1, i -= Fib[x] else n .= 0 } Out .= (n ? LTrim(n, "0") : 0) "`n" } MsgBox, % Out
NStepSequence(v1, v2, n, k) {
a := [v1, v2]
Loop, % k - 2 { a[j := A_Index + 2] := 0 Loop, % j < n + 2 ? j - 1 : n a[j] += a[j - A_Index] } return, a }</lang>NStepSequence() Output:
0: 0 1: 1 2: 10 3: 100 4: 101 5: 1000 6: 1001 7: 1010 8: 10000 9: 10001 10: 10010 11: 10100 12: 10101 13: 100000 14: 100001 15: 100010 16: 100100 17: 100101 18: 101000 19: 101001 20: 101010
AutoIt
<lang autoit> For $i = 0 To 20 ConsoleWrite($i &": "& Zeckendorf($i)&@CRLF) Next
Func Zeckendorf($int, $Fibarray = "") If Not IsArray($Fibarray) Then $Fibarray = Fibonacci($int) Local $ret = "" For $i = UBound($Fibarray) - 1 To 1 Step -1 If $Fibarray[$i] > $int And $ret = "" Then ContinueLoop ; dont use Leading Zeros If $Fibarray[$i] > $int Then $ret &= "0" Else If StringRight($ret, 1) <> "1" Then $ret &= "1" $int -= $Fibarray[$i] Else $ret &= "0" EndIf EndIf Next If $ret = "" Then $ret = "0" Return $ret EndFunc ;==>Zeckendorf
Func Fibonacci($max) $AList = ObjCreate("System.Collections.ArrayList") $AList.add("0") $current = 0 While True If $current > 1 Then $count = $AList.Count $current = $AList.Item($count - 1) $current = $current + $AList.Item($count - 2) Else $current += 1 EndIf $AList.add($current) If $current > $max Then ExitLoop WEnd $Array = $AList.ToArray Return $Array EndFunc ;==>Fibonacci </lang> Output:
0: 0 1: 1 2: 10 3: 100 4: 101 5: 1000 6: 1001 7: 1010 8: 10000 9: 10001 10: 10010 11: 10100 12: 10101 13: 100000 14: 100001 15: 100010 16: 100100 17: 100101 18: 101000 19: 101001 20: 101010
BBC BASIC
<lang bbcbasic> FOR n% = 0 TO 20
PRINT n% RIGHT$(" " + FNzeckendorf(n%), 8) NEXT PRINT '"Checking numbers up to 10000..." FOR n% = 21 TO 10000 IF INSTR(FNzeckendorf(n%), "11") STOP NEXT PRINT "No Zeckendorf numbers contain consecutive 1's" END DEF FNzeckendorf(n%) LOCAL i%, o$, fib%() : DIM fib%(45) fib%(0) = 1 : fib%(1) = 1 : i% = 1 REPEAT i% += 1 fib%(i%) = fib%(i%-1) + fib%(i%-2) UNTIL fib%(i%) > n% REPEAT i% -= 1 IF n% >= fib%(i%) THEN o$ += "1" n% -= fib%(i%) ELSE o$ += "0" ENDIF UNTIL i% = 1 = o$</lang>
Output:
0 0 1 1 2 10 3 100 4 101 5 1000 6 1001 7 1010 8 10000 9 10001 10 10010 11 10100 12 10101 13 100000 14 100001 15 100010 16 100100 17 100101 18 101000 19 101001 20 101010 Checking numbers up to 10000... No Zeckendorf numbers contain consecutive 1's
Befunge
The first number on the stack, 45*, specifies the range of values to display. However, the algorithm depends on a hardcoded list of Fibonacci values (currently just 10) so the theoretical maximum is 143. It's also constrained by the range of a Befunge data cell, which on many implementations will be as low as 127.
<lang befunge>45*83p0>:::.0`"0"v v53210p 39+!:,,9+< >858+37 *66g"7Y":v >3g`#@_^ v\g39$< ^8:+1,+5_5<>-:0\`| v:-\g39_^#:<*:p39< >0\`:!"0"+#^ ,#$_^</lang>
- Output:
0 0 1 1 2 10 3 100 4 101 5 1000 6 1001 7 1010 8 10000 9 10001 10 10010 11 10100 12 10101 13 100000 14 100001 15 100010 16 100100 17 100101 18 101000 19 101001 20 101010
C
<lang c>#include <stdio.h>
typedef unsigned long long u64;
- define FIB_INVALID (~(u64)0)
u64 fib[] = { 1, 2, 3, 5, 8, 13, 21, 34, 55, 89, 144, 233, 377, 610, 987, 1597, 2584, 4181, 6765, 10946, 17711, 28657, 46368, 75025, 121393, 196418, 317811, 514229, 832040, 1346269, 2178309, 3524578, 5702887, 9227465, 14930352, 24157817, 39088169, 63245986, 102334155, 165580141, 267914296, 433494437, 701408733, 1134903170, 1836311903, 2971215073ULL, 4807526976ULL, 7778742049ULL, 12586269025ULL, 20365011074ULL, 32951280099ULL, 53316291173ULL, 86267571272ULL, 139583862445ULL, 225851433717ULL, 365435296162ULL, 591286729879ULL, 956722026041ULL, 1548008755920ULL, 2504730781961ULL, 4052739537881ULL, 6557470319842ULL, 10610209857723ULL, 17167680177565ULL,
27777890035288ULL // this 65-th one is for range check };
u64 fibbinary(u64 n) { if (n >= fib[64]) return FIB_INVALID;
u64 ret = 0; int i; for (i = 64; i--; ) if (n >= fib[i]) { ret |= 1ULL << i; n -= fib[i]; }
return ret; }
void bprint(u64 n, int width) { if (width > 64) width = 64;
u64 b; for (b = 1ULL << (width - 1); b; b >>= 1) putchar(b == 1 && !n ? '0' : b > n ? ' ' : b & n ? '1' : '0'); putchar('\n'); }
int main(void) { int i;
for (i = 0; i <= 20; i++) printf("%2d:", i), bprint(fibbinary(i), 8);
return 0; }</lang>
- Output:
0: 0 1: 1 2: 10 3: 100 4: 101 5: 1000 6: 1001 7: 1010 8: 10000 9: 10001 10: 10010 11: 10100 12: 10101 13: 100000 14: 100001 15: 100010 16: 100100 17: 100101 18: 101000 19: 101001 20: 101010
C++
Using a C++11 User Defined Literal
see Here for a further example using this class. <lang cpp> // For a class N which implements Zeckendorf numbers: // I define an increment operation ++() // I define a comparison operation <=(other N) // Nigel Galloway October 22nd., 2012
- include <iostream>
class N { private:
int dVal = 0, dLen;
public:
N(char const* x = "0"){ int i = 0, q = 1; for (; x[i] > 0; i++); for (dLen = --i/2; i >= 0; i--) { dVal+=(x[i]-48)*q; q*=2; }} const N& operator++() { for (int i = 0;;i++) { if (dLen < i) dLen = i; switch ((dVal >> (i*2)) & 3) { case 0: dVal += (1 << (i*2)); return *this; case 1: dVal += (1 << (i*2)); if (((dVal >> ((i+1)*2)) & 1) != 1) return *this; case 2: dVal &= ~(3 << (i*2)); }}} const bool operator<=(const N& other) const {return dVal <= other.dVal;} friend std::ostream& operator<<(std::ostream&, const N&);
}; N operator "" N(char const* x) {return N(x);} std::ostream &operator<<(std::ostream &os, const N &G) {
const static std::string dig[] {"00","01","10"}, dig1[] {"","1","10"}; if (G.dVal == 0) return os << "0"; os << dig1[(G.dVal >> (G.dLen*2)) & 3]; for (int i = G.dLen-1; i >= 0; i--) os << dig[(G.dVal >> (i*2)) & 3]; return os;
} </lang> I may now write: <lang cpp> int main(void) { //for (N G; G <= 101010N; ++G) std::cout << G << std::endl; // from zero to 101010M
for (N G(101N); G <= 101010N; ++G) std::cout << G << std::endl; // from 101N to 101010N return 0;
} </lang> Which produces:
- Output:
101 1000 1001 1010 10000 10001 10010 10100 10101 100000 100001 100010 100100 100101 101000 101001 101010
C#
<lang csharp> using System; using System.Collections.Generic; using System.Linq; using System.Text;
namespace Zeckendorf {
class Program { private static uint Fibonacci(uint n) { if (n < 2) { return n; } else { return Fibonacci(n - 1) + Fibonacci(n - 2); } }
private static string Zeckendorf(uint num) { IList<uint> fibonacciNumbers = new List<uint>(); uint fibPosition = 2;
uint currentFibonaciNum = Fibonacci(fibPosition);
do { fibonacciNumbers.Add(currentFibonaciNum); currentFibonaciNum = Fibonacci(++fibPosition); } while (currentFibonaciNum <= num);
uint temp = num; StringBuilder output = new StringBuilder();
foreach (uint item in fibonacciNumbers.Reverse()) { if (item <= temp) { output.Append("1"); temp -= item; } else { output.Append("0"); } }
return output.ToString(); }
static void Main(string[] args) { for (uint i = 1; i <= 20; i++) { string zeckendorfRepresentation = Zeckendorf(i); Console.WriteLine(string.Format("{0} : {1}", i, zeckendorfRepresentation)); }
Console.ReadKey(); } }
} </lang>
- Output:
1 : 1 2 : 10 3 : 100 4 : 101 5 : 1000 6 : 1001 7 : 1010 8 : 10000 9 : 10001 10 : 10010 11 : 10100 12 : 10101 13 : 100000 14 : 100001 15 : 100010 16 : 100100 17 : 100101 18 : 101000 19 : 101001 20 : 101010
Clojure
<lang clojure>(def fibs (lazy-cat [1 1] (map + fibs (rest fibs))))
(defn z [n]
(if (zero? n) "0" (let [ps (->> fibs (take-while #(<= % n)) rest reverse) fz (fn [[s n] p] (if (>= n p) [(conj s 1) (- n p)] [(conj s 0) n]))] (->> ps (reduce fz [[] n]) first (apply str)))))
(doseq [n (range 0 21)] (println n (z n)))</lang>
Common Lisp
Common Lisp's arbitrary precision integers should handle any positive input: <lang lisp>(defun zeckendorf (n)
"returns zeckendorf integer of n (see OEIS A003714)" (let ((fib '(2 1)))
;; extend Fibonacci sequence long enough (loop while (<= (car fib) n) do (push (+ (car fib) (cadr fib)) fib)) (loop with r = 0 for f in fib do (setf r (* 2 r)) (when (>= n f) (setf n (- n f)) (incf r)) finally (return r))))
- task requirement
(loop for i from 0 to 20 do
(format t "~2D: ~2R~%" i (zeckendorf i)))</lang>
<lang lisp>
- Print Zeckendorf numbers upto 20.
- I have implemented this as a state machine.
- Nigel Galloway - October 13th., 2012
(let ((fibz '(13 8 5 3 2 1))) (dotimes (G 21) (progn (format t "~S is " G)
(let ((z 0) (ng G)) (dolist (N fibz) (if (> z 1) (progn (setq z 1) (format t "~S" 0)) (if (>= ng N) (progn (setq z 2) (setq ng (- ng N)) (format t "~S" 1)) (if (= z 1) (format t "~S" 0))))) (if (= z 0) (format t "~S~%" 0) (format t "~%"))))))
</lang>
- Output:
0 is 0 1 is 1 2 is 10 3 is 100 4 is 101 5 is 1000 6 is 1001 7 is 1010 8 is 10000 9 is 10001 10 is 10010 11 is 10100 12 is 10101 13 is 100000 14 is 100001 15 is 100010 16 is 100100 17 is 100101 18 is 101000 19 is 101001 20 is 101010
D
<lang d>import std.stdio, std.range, std.algorithm, std.functional;
void main() {
size_t .max .iota .filter!q{ !(a & (a >> 1)) } .take(21) .binaryReverseArgs!writefln("%(%b\n%)");
}</lang>
- Output:
0 1 10 100 101 1000 1001 1010 10000 10001 10010 10100 10101 100000 100001 100010 100100 100101 101000 101001 101010
<lang d>import std.stdio, std.typecons;
int zeckendorf(in int n) pure nothrow {
Tuple!(int,"remaining", int,"set") zr(in int fib0, in int fib1, in int n, in uint bit) pure nothrow { if (fib1 > n) return typeof(return)(n, 0); auto rs = zr(fib1, fib0 + fib1, n, bit + 1); if (fib1 <= rs.remaining) { rs.set |= 1 << bit; rs.remaining -= fib1; } return rs; }
return zr(1, 1, n, 0)[1];
}
void main() {
foreach (i; 0 .. 21) writefln("%2d: %6b", i, zeckendorf(i));
}</lang>
- Output:
0: 0 1: 1 2: 10 3: 100 4: 101 5: 1000 6: 1001 7: 1010 8: 10000 9: 10001 10: 10010 11: 10100 12: 10101 13: 100000 14: 100001 15: 100010 16: 100100 17: 100101 18: 101000 19: 101001 20: 101010
(Same output.) <lang d>import std.stdio, std.algorithm, std.range;
string zeckendorf(size_t n) {
if (n == 0) return "0"; auto fibs = recurrence!q{a[n - 1] + a[n - 2]}(1, 2);
string result; foreach_reverse (immutable f; fibs.until!(x => x > n).array) { result ~= (f <= n) ? '1' : '0'; if (f <= n) n -= f; }
return result;
}
void main() {
foreach (immutable i; 0 .. 21) writefln("%2d: %6s", i, i.zeckendorf);
}</lang>
EchoLisp
We analytically find the first fibonacci(i) >= n, using the formula i = log((n* Φ) + 0.5) / log(Φ) . <lang scheme>
- special fib's starting with 1 2 3 5 ...
(define (fibonacci n)
(+ (fibonacci (1- n)) (fibonacci (- n 2))))
(remember 'fibonacci #(1 2))
(define-constant Φ (// (1+ (sqrt 5)) 2)) (define-constant logΦ (log Φ))
- find i
- fib(i) >= n
(define (iFib n)
(floor (// (log (+ (* n Φ) 0.5)) logΦ)))
- left trim zeroes
(string-delimiter "") (define (zeck->string digits)
(if (!= 0 (first digits)) (string-join digits "") (zeck->string (rest digits))))
(define (Zeck n)
(cond (( < n 0) "no negative zeck") ((inexact? n) "no floating zeck") ((zero? n) "0") (else (zeck->string (for/list ((s (reverse (take fibonacci (iFib n))))) (if ( > s n) 0 (begin (-= n s) 1 )))))))
</lang>
- Output:
(take Zeck 21) → (0 1 10 100 101 1000 1001 1010 10000 10001 10010 10100 10101 100000 100001 100010 100100 100101 101000 101001 101010) (Zeck 1000000000) → 1010000100100001010101000001000101000101001
Elena
ELENA 3.4 : <lang elena>import system'routines. import system'collections. import system'text. import extensions.
extension op {
fibonacci [ if (self < 2) [ ^ self ]; [ ^ (self - 1) fibonacci + (self - 2) fibonacci. ]. ] zeckendorf [ var fibonacciNumbers := List<IntNumber>(10). int num := self. int fibPosition := 2. int currentFibonaciNum := fibPosition fibonacci. while (currentFibonaciNum <= num) [ fibonacciNumbers append:currentFibonaciNum. fibPosition := fibPosition + 1. currentFibonaciNum := fibPosition fibonacci. ]. auto output := TextBuilder new. int temp := num. fibonacciNumbers sequenceReverse; forEach(:item) [ if (item <= temp) [ output << "1". temp := temp - item. ]; [ output << "0". ]. ]. ^ output literal. ]
}
public program [
1 to:20 do(:i) [ var zeckendorfRepresentation := i zeckendorf. console printFormatted("{0} : {1}",i,i zeckendorf); writeLine. ]. console readChar.
]</lang>
- Output:
1 : 1 2 : 10 3 : 100 4 : 101 5 : 1000 6 : 1001 7 : 1010 8 : 10000 9 : 10001 10 : 10010 11 : 10100 12 : 10101 13 : 100000 14 : 100001 15 : 100010 16 : 100100 17 : 100101 18 : 101000 19 : 101001 20 : 101010
Elixir
Stream generator: <lang elixir>defmodule Zeckendorf do
def number do Stream.unfold(0, fn n -> zn_loop(n) end) end defp zn_loop(n) do bin = Integer.to_string(n, 2) if String.match?(bin, ~r/11/), do: zn_loop(n+1), else: {bin, n+1} end
end
Zeckendorf.number |> Enum.take(21) |> Enum.with_index |> Enum.each(fn {zn, i} -> IO.puts "#{i}: #{zn}" end)</lang>
- Output:
0: 0 1: 1 2: 10 3: 100 4: 101 5: 1000 6: 1001 7: 1010 8: 10000 9: 10001 10: 10010 11: 10100 12: 10101 13: 100000 14: 100001 15: 100010 16: 100100 17: 100101 18: 101000 19: 101001 20: 101010
Fibonacci numbers: <lang elixir>defmodule Zeckendorf do
def number(n) do fib_loop(n, [2,1]) |> Enum.reduce({"",n}, fn f,{dig,i} -> if f <= i, do: {dig<>"1", i-f}, else: {dig<>"0", i} end) |> elem(0) |> String.to_integer end defp fib_loop(n, fib) when n < hd(fib), do: fib defp fib_loop(n, [a,b|_]=fib), do: fib_loop(n, [a+b | fib])
end
for i <- 0..20, do: IO.puts "#{i}: #{Zeckendorf.number(i)}"</lang> same output
F#
<lang fsharp>let fib = Seq.unfold (fun (x, y) -> Some(x, (y, x + y))) (1,2)
let zeckendorf n =
if n = 0 then ["0"] else let folder k state = let (n, z) = (fst state), (snd state) if n >= k then (n - k, "1" :: z) else (n, "0" :: z) let fb = fib |> Seq.takeWhile (fun i -> i<=n) |> Seq.toList snd (List.foldBack folder fb (n, [])) |> List.rev
for i in 0 .. 20 do printfn "%2d: %8s" i (String.concat "" (zeckendorf i))</lang>
- Output:
0: 0 1: 1 2: 10 3: 100 4: 101 5: 1000 6: 1001 7: 1010 8: 10000 9: 10001 10: 10010 11: 10100 12: 10101 13: 100000 14: 100001 15: 100010 16: 100100 17: 100101 18: 101000 19: 101001 20: 101010
Factor
<lang factor>USING: formatting kernel locals make math sequences ;
- fib<= ( n -- seq )
1 2 [ [ dup n <= ] [ 2dup + [ , ] 2dip ] while drop , ] { } make ;
- zeck ( n -- str )
0 :> s! n fib<= <reversed> [ dup s + n <= [ s + s! 49 ] [ drop 48 ] if ] "" map-as ;
21 <iota> [ dup zeck "%2d: %6s\n" printf ] each</lang>
- Output:
0: 0 1: 1 2: 10 3: 100 4: 101 5: 1000 6: 1001 7: 1010 8: 10000 9: 10001 10: 10010 11: 10100 12: 10101 13: 100000 14: 100001 15: 100010 16: 100100 17: 100101 18: 101000 19: 101001 20: 101010
Forth
<lang forth>: fib<= ( n -- n )
>r 0 1 BEGIN dup r@ <= WHILE tuck + REPEAT drop rdrop ;
- z. ( n -- )
dup fib<= dup . - BEGIN ?dup WHILE dup fib<= dup [char] + emit space . - REPEAT ;
- tab 9 emit ;
- zeckendorf ( -- )
21 0 DO cr i 2 .r tab i z. LOOP ;</lang>
- Output:
zeckendorf 0 0 1 1 2 2 3 3 4 3 + 1 5 5 6 5 + 1 7 5 + 2 8 8 9 8 + 1 10 8 + 2 11 8 + 3 12 8 + 3 + 1 13 13 14 13 + 1 15 13 + 2 16 13 + 3 17 13 + 3 + 1 18 13 + 5 19 13 + 5 + 1 20 13 + 5 + 2 ok
Fortran
The simplest representation of a number in the Zeckendorf manner is as a sequence of digits, such as are used in multi-precision arithmetic, and for this an array of integers will do. Rather excessively, as only two states are required and the default integer style is usually thirty-two bits these days. Some compilers allow the specification of one-byte integers, as in INTEGER*1 D(0:ZLAST)
so that would be only an eight-fold excess. One could escalate to fiddling with individual bits within a number (as is done in Extensible_prime_generator#Fortran) and a 16-bit integer would be adequate for the specified tests, but Fortran syntax has not been extended to offer simple methods for manipulating bits such as D(7:7)
to obtain the seventh bit of D. Instead one might use special library routines as supplied by F90 such as IBITS(D,7,1)
for the same effect, though possibly at a cost in code size and execution time. Less storage may be saved through cramming bits than is consumed by the code needed to extract individual bits. Such values could then be displayed using theB
format code. However, the source code would now be littered with the details of bit access rather than the form of the Zeckendorf procedure.
An alternative lies in noting that only the sequences 00, 01, and 10 can appear (because 11 is unnecessary; see below), so a base three scheme could be used to represent the three such pairs of bits. But this still contains redundancies. Suppose a 01 value is somewhere in the sequence: then it may be followed only by 00 or 01, and likewise 10 be preceded only by 10 or 00; just two values, not three. Perhaps a still more cunning compaction scheme could be devised to take advantage of these details, or some other scheme concocted. For simplicity, no compaction will be attempted so the states of 0 and 1 will be represented by a simple integer devoted to that bit.
The conversion scheme requires the values of the Fibonacci sequence, except not quite: the Fibonacci sequence starts F0 = 0, F1 = 1, F2 = 1, F3 = 2, F4 = 3, etc. but what is wanted is to start with the second 1, so F1 = 1, F2 = 2, F3 = 3, F4 = 5, etc. so this sequence has been named the Fib1nacci sequence to replace conceptual dissonance with lexical dissonance, and similarly with array F1B
instead of FIB
. Initial investigations show that F45 is the last before a thirty-two bit two's complement integer is overflowed, though systems offering INTEGER*8 could be pushed further.
Initialising this table in array F1B
could be done via specifying the relevant values (computed separately, even by hand) or by some banal initialisation loop that would be executed on the first invocation of any of the routines requiring those values, a tedious and annoying rigmarole to organise. More interesting are the facilities offered by the PARAMETER statement (introduced with F77), with the further possibility that constants, so defined, would be held safe from accidental change, nor would there be initialising code to execute at run time. Alas, the obvious approach (commented out in the source) using array F1B
is rejected by the F90 compiler even though it does allow a value to be defined in terms of other defined values, as is demonstrated by the horde of simple variables following. Despite being a multi-pass compiler, the dependencies will not be unravelled if the statements appear out of order. Fortran does not include a standard pre-processor stage, unlike say pl/i where it is built-in to the language and uses much the same syntax as normal pl/i statements, so loops, IF-tests and so forth are available. By such means, the upper limit of 45 could be determined, the initial values calculated, and the array be defined with initial values, all at compile time.
By declaring the horde of simple names to have the PRIVATE attribute, they will not litter the name space of routines invoking the module, but alas, they will still occupy their own storage space. Another possibility would be to use the EQUIVALENCE statement to have them placed within array F1B
, but alas, as noted in 15_Puzzle_Game#Fortran, the compiler will not countenance PARAMETER statements for names engaged in such misbehviour. A pity.
Still another possibility would be to take advantage of the formula for calculating the values of the Fibonacci series directly (with careful attention to the offsets needed for the Fib1nacci sequence), but this formula is rather intimidating: <lang Fortran>F(N) = ((1 + SQRT(5))**N - (1 - SQRT(5))**N)/(SQRT(5)*2**N)</lang> It can easily be coded as a Fortran function (and would have to be double precision because 32-bit floating-point arithmetic is not accurate enough for integer constants approaching 32 bits), but alas, the compiler does not allow itself to take the risk of invoking a user-written function in a PARAMETER statement, even if the compiler had itself compiled it. For, who knows what it might do?
Given the array F1B
the conversion from an integer to Zeckendorf digit sequence starts from the high-order end to find the highest F1B
value not exceeding the number. There is a formula for this mentioned in the EchoLisp section, but it too is intimidating and the rounding of its result would also require checking. Rather than a linear search, a binary chop could be used, though at the cost of additional code. The location of the high-order digit is recorded on general principles, it being useful in formatting output for example. This requires a "first-time" test within the loop, that could be avoided if the conversion were to be done in two stages.
The special feature of the conversion lies in noting that F1B(n + 1) = F1B(n) + F1B(n - 1), the defining feature of the Fibonacci sequence. Thus, when a 1-bit is found (say it is bit n), the next bit down must be a 0 and so the test for it may be skipped, by incrementing L
This is because if it were not 0, then the bit above (bit n + 1) would have been turned on in the previous stage instead. Because of this adjustment, the controlling loop cannot be DO L = ZLAST,1,-1
to step down the entries in array F1B
because modifications to the index variable of a DO-loop, if not rejected out-of-hand by the compiler, may have no effect on the execution of the loop. This is because the execution of a DO-loop is often controlled by an "iteration count", calculated on entry to the DO-loop, which is thereby unaffected by changes to the index variable, or indeed to the bounds and step size of the loop. Other implementations of a DO-loop will offer other behaviour. There being no equivalent in Fortran of Repeat ... until ... ; (whereby the test is at the end, and there is no initial test), a GO TO
appears...
The source uses F90 for its MODULE facility, in particular having array F1B
available without having to mess about with additional parameters or COMMON statements. This also enables the specification of arrays with a lower bound other than one, which makes it easy to define the digit arrays to have a current length, stored in element zero. This sort of "string" facility is often restricted only to strings of characters, but the notion "string of <type>" is often useful. If in routines declared within a MODULE the size of an array parameter is declared via :
there are secret additional parameters defining its size, accessible via special functions such as UBOUND
so there is no need for an explicit parameter doing so as would be the case prior to F90. With F90 it is also possible to define a compound data type for the digit sequence, but a simple array seems more flexible.
The pleasing name, "MODULE ZECKENDORF ARITHMETIC" causes some odd behaviour, even though Fortran source normally involves spaces having no significance outside text literals.<lang Fortran> MODULE ZECKENDORF ARITHMETIC !Using the Fibonacci series, rather than powers of some base.
INTEGER ZLAST !The standard 32-bit two's complement integers PARAMETER (ZLAST = 45) !only get so far, just as there's a limit to the highest power. INTEGER F1B(ZLAST) !I want the Fibonacci series, and, starting with its second one.
c PARAMETER (F1B = (/1,2, !But alas, the compiler doesn't allow c 3 F1B(1) + F1B(2), !for this sort of carpet-unrolling c 4 F1B(2) + F1B(3), !initialisation sequence.
INTEGER,PRIVATE:: F01,F02,F03,F04,F05,F06,F07,F08,F09,F10, !So, not bothering with F00, 1 F11,F12,F13,F14,F15,F16,F17,F18,F19,F20, !Prepare a horde of simple names, 2 F21,F22,F23,F24,F25,F26,F27,F28,F29,F30, !which can be initialised 3 F31,F32,F33,F34,F35,F36,F37,F38,F39,F40, !in a certain way, 4 F41,F42,F43,F44,F45 !without scaring the compiler. PARAMETER (F01 = 1, F02 = 2, F03 = F02 + F01, F04 = F03 + F02, !Thusly. 1 F05=F04+F03,F06=F05+F04,F07=F06+F05,F08=F07+F06,F09=F08+F07, !Typing all this 2 F10=F09+F08,F11=F10+F09,F12=F11+F10,F13=F12+F11,F14=F13+F12, !is an invitation 3 F15=F14+F13,F16=F15+F14,F17=F16+F15,F18=F17+F16,F19=F18+F17, !for mistypes. 4 F20=F19+F18,F21=F20+F19,F22=F21+F20,F23=F22+F21,F24=F23+F22, !So a regular layout 5 F25=F24+F23,F26=F25+F24,F27=F26+F25,F28=F27+F26,F29=F28+F27, !helps a little. 6 F30=F29+F28,F31=F30+F29,F32=F31+F30,F33=F32+F31,F34=F33+F32, !Otherwise, 7 F35=F34+F33,F36=F35+F34,F37=F36+F35,F38=F37+F36,F39=F38+F37, !devise a prog. 8 F40=F39+F38,F41=F40+F39,F42=F41+F40,F43=F42+F41,F44=F43+F42, !to generate these texts... 9 F45=F44+F43) !The next is 2971215073. Too big for 32-bit two's complement integers. PARAMETER (F1B = (/F01,F02,F03,F04,F05,F06,F07,F08,F09,F10, !And now, 1 F11, F12, F13, F14, F15, F16, F17, F18, F19, F20, !Here is the desired 2 F21, F22, F23, F24, F25, F26, F27, F28, F29, F30, !array of constants. 3 F31, F32, F33, F34, F35, F36, F37, F38, F39, F40, !And as such, possibly 4 F41, F42, F43, F44, F45/)) !protected from alteration. CONTAINS !After all that, here we go. SUBROUTINE ZECK(N,D) !Convert N to a "Zeckendorf" digit sequence.
Counts upwards from digit one. D(i) ~ F1B(i). D(0) fingers the high-order digit.
INTEGER N !The normal number, in the computer's base. INTEGER D(0:) !The digits, to be determined. INTEGER R !The remnant. INTEGER L !A finger, similar to the power of the base. IF (N.LT.0) STOP "ZECK! No negative numbers!" !I'm not thinking about them. R = N !Grab a copy that I can mess with. D = 0 !Scrub the lot in one go. L = ZLAST !As if starting with BASE**MAX, rather than BASE**0. 10 IF (R.GE.F1B(L)) THEN !Has the remnant sufficient for this digit? R = R - F1B(L) !Yes! Remove that amount. IF (D(0).EQ.0) THEN !Is this the first non-zero digit? IF (L.GT.UBOUND(D,DIM=1)) STOP "ZECK! Not enough digits!" !Yes. D(0) = L !Remember the location of the high-order digit. END IF !Two loops instead, to avoid repeated testing? D(L) = 1 !Place the digit, knowing a place awaits. L = L - 1 !Never need a ...11... sequence because F1B(i) + F1B(i+1) = F1B(i+2). END IF !So much for that digit "power". L = L - 1 !Down a digit. IF (L.GT.0 .AND. R.GT.0) GO TO 10 !Are we there yet? IF (N.EQ.0) D(0) = 1 !Zero has one digit. END SUBROUTINE ZECK !That was fun.
INTEGER FUNCTION ZECKN(D) !Converts a "Zeckendorf" digit sequence to a number. INTEGER D(0:) !The digits. D(0) fingers the high-order digit. IF (D(0).LE.0) STOP "ZECKN! Empty number!" !General paranoia. IF (D(0).GT.ZLAST) STOP "ZECKN! Oversize number!" !I hate array bound hiccoughs. ZECKN = SUM(D(1:D(0))*F1B(1:D(0))) !This is what positional notation means. IF (ZECKN.LT.0) STOP "ZECKN! Integer overflow!" !Oh for IF OVERFLOW as in First Fortran. END FUNCTION ZECKN !Overflows by a small amount will produce a negative number. END MODULE ZECKENDORF ARITHMETIC !Odd stuff.
PROGRAM POKE USE ZECKENDORF ARITHMETIC !Please. INTEGER ZD(0:ZLAST) !A scratchpad. INTEGER I,J,W CHARACTER*1 DIGIT(0:1) !Assistance for the output. PARAMETER (DIGIT = (/"0","1"/), W = 6) !This field width suffices.
c WRITE (6,*) F1B c WRITE (6,*) INT8(F1B(44)) + INT8(F1B(45))
WRITE (6,1) F1B(1:4),ZLAST,ZLAST,F1B(ZLAST),HUGE(I) !Show some provenance. 1 FORMAT ("Converts integers to their Zeckendorf digit string " 1 "using the Fib1nacci sequence (",4(I0,","), 2 " ...) as the equivalent of powers."/ 3 "At most, ",I0," digits because Fib1nacci(",I0,") = ",I0, 4 " and the integer limit is ",I0,".",//," N ZN") !Ends with a heading.
DO I = 0,20 !Step through the specified range. CALL ZECK(I,ZD) !Convert I to ZD.
c WRITE (6,2) I,ZD(ZD(0):1:-1) !Show digits from high-order to low. c 2 FORMAT (I3,1X,66I1) !Or, WRITE (6,2) I,(ZD(J), J = ZD(0),1,-1)
WRITE (6,3) I,(" ",J = ZD(0) + 1,W),DIGIT(ZD(ZD(0):1:-1)) !Right-aligned in field width W. 3 FORMAT (I3,1X,66A1) !The digits appear as characters. IF (I.NE.ZECKN(ZD)) STOP "Huh?" !Should never happen... END DO !On to the next.
END</lang>
Output: shown aligned right for a more regular table. Producing leading spaces or digits required a conversion from a numerical digit to a character digit, so that all the output could use the A
format code.
Converts integers to their Zeckendorf digit string using the Fib1nacci sequence (1,2,3,5, ...) as the equivalent of powers. At most, 45 digits because Fib1nacci(45) = 1836311903 and the integer limit is 2147483647. N ZN 0 0 1 1 2 10 3 100 4 101 5 1000 6 1001 7 1010 8 10000 9 10001 10 10010 11 10100 12 10101 13 100000 14 100001 15 100010 16 100100 17 100101 18 101000 19 101001 20 101010
FreeBASIC
<lang freebasic>' version 17-10-2016 ' compile with: fbc -s console
- Define max 92 ' max for Fibonacci number
Dim Shared As ULongInt fib(max)
fib(0) = 1 fib(1) = 1
For x As Integer = 2 To max
fib(x) = fib(x-1) + fib(x-2)
Next
Function num2zeck(n As Integer) As String
If n < 0 Then
Print "Error: no negative numbers allowed" Beep : Sleep 5000,1 : End
End If
If n < 2 Then Return Str(n)
Dim As String zeckendorf For x As Integer = max To 1 Step -1 If fib(x) <= n Then zeckendorf = zeckendorf + "1" n = n - fib(x) Else zeckendorf = zeckendorf + "0" End If Next
return LTrim(zeckendorf, "0") ' get rid of leading zeroes
End Function
' ------=< MAIN >=------
Dim As Integer x, e Dim As String zeckendorf Print "number zeckendorf"
For x = 0 To 200000
zeckendorf = num2zeck(x) If x <= 20 Then Print x, zeckendorf ' check for two consecutive Fibonacci numbers If InStr(zeckendorf, "11") <> 0 Then Print " Error: two consecutive Fibonacci numbers "; x, zeckendorf e = e +1 End If
Next
Print If e = 0 Then
Print " No Zeckendorf numbers with two consecutive Fibonacci numbers found"
Else
Print e; " error(s) found"
End If
' empty keyboard buffer
While Inkey <> "" : Wend
Print : Print "hit any key to end program"
Sleep
End</lang>
- Output:
number zeckendorf 0 0 1 1 2 10 3 100 4 101 5 1000 6 1001 7 1010 8 10000 9 10001 10 10010 11 10100 12 10101 13 100000 14 100001 15 100010 16 100100 17 100101 18 101000 19 101001 20 101010 No Zeckendorf numbers with two consecutive Fibonacci numbers found
Go
<lang go>package main
import "fmt"
func main() {
for i := 0; i <= 20; i++ { fmt.Printf("%2d %7b\n", i, zeckendorf(i)) }
}
func zeckendorf(n int) int {
// initial arguments of fib0 = 1 and fib1 = 1 will produce // the Fibonacci sequence {1, 2, 3,..} on the stack as successive // values of fib1. _, set := zr(1, 1, n, 0) return set
}
func zr(fib0, fib1, n int, bit uint) (remaining, set int) {
if fib1 > n { return n, 0 } // recurse. // construct sequence on the way in, construct ZR on the way out. remaining, set = zr(fib1, fib0+fib1, n, bit+1) if fib1 <= remaining { set |= 1 << bit remaining -= fib1 } return
}</lang>
- Output:
0 0 1 1 2 10 3 100 4 101 5 1000 6 1001 7 1010 8 10000 9 10001 10 10010 11 10100 12 10101 13 100000 14 100001 15 100010 16 100100 17 100101 18 101000 19 101001 20 101010
Haskell
Using "no consecutive 1s" rule: <lang haskell>import Data.Bits import Numeric
zeckendorf = map b $ filter ones [0..] where ones :: Int -> Bool ones x = 0 == x .&. (x `shiftR` 1) b x = showIntAtBase 2 ("01"!!) x ""
main = mapM_ putStrLn $ take 21 zeckendorf</lang> which is the same as <lang haskell>zeckendorf = "0":"1":[s++[d] | s <- tail zeckendorf, d <- "01", last s /= '1' || d /= '1']
main = mapM putStrLn $ take 21 zeckendorf</lang> or a different way to generate the sequence: <lang haskell>import Numeric
fib = 1 : 1 : zipWith (+) fib (tail fib) pow2 = iterate (2*) 1
zeckendorf = map b z where z = 0:concat (zipWith f fib pow2) f x y = map (y+) (take x z) b x = showIntAtBase 2 ("01"!!) x ""
main = mapM_ putStrLn $ take 21 zeckendorf</lang>
Creating a string for an individual number: <lang haskell>import Data.List (mapAccumL)
fib :: [Int] fib = 1 : 2 : zipWith (+) fib (tail fib)
zeckendorf :: Int -> String zeckendorf 0 = "0" zeckendorf n = snd $ mapAccumL f n $ reverse $ takeWhile (<= n) fib
where f n x | n < x = (n, '0') | otherwise = (n - x, '1')
main :: IO () main = (putStrLn . unlines) $ zeckendorf <$> [0 .. 20]</lang>
- Output:
0 1 10 100 101 1000 1001 1010 10000 10001 10010 10100 10101 100000 100001 100010 100100 100101 101000 101001 101010
J
Please enjoy our Zeckendorf essay. <lang J> fib=: 3 : 0 " 0
mp=. +/ .* {.{: mp/ mp~^:(I.|.#:y) 2 2$0 1 1 1x
)
phi=: -:1+%:5
fi =: 3 : 'n - y<fib n=. 0>.(1=y)-~>.(phi^.%:5)+phi^.y'
fsum=: 3 : 0
z=. 0$r=. y while. 3<r do. m=. fib fi r z=. z,m r=. r-m end. z,r$~(*r)+.0=y
)
Filter=: (#~`)(`:6)
' '&~:Filter@:":@:#:@:#.@:((|. fib 2+i.8) e. fsum)&.>i.3 7 ┌──────┬──────┬──────┬──────┬──────┬──────┬──────┐ │0 │1 │10 │100 │101 │1000 │1001 │ ├──────┼──────┼──────┼──────┼──────┼──────┼──────┤ │1010 │10000 │10001 │10010 │10100 │10101 │100000│ ├──────┼──────┼──────┼──────┼──────┼──────┼──────┤ │100001│100010│100100│100101│101000│101001│101010│ └──────┴──────┴──────┴──────┴──────┴──────┴──────┘ </lang>
Explanation:
fsum
finds the canonical list of fibonacci terms which sum to its argument.
fib
finds the nth fibonacci term of the fibonacci sequence. This would be 0 1 1 2 3 5 8 13 21 34 55 89 ... but we ignore the first two values of that sequence for the purpose of this exercise.
(|. fib 2+i.8)
is 34 21 13 8 5 3 2 1
. You can think of an eight bit Zeckendorf number such as 101010
as representing the inner product of its digits with (|. fib 2+i.8)
. Thus, we can find the relevant Zeckendorf bits by finding which which members of that sequence are in the result of fsum
The rest is just formatting. (We convert from binary list to integer and then back to binary list, to eliminate leading zeros from the list. Then we convert to text and remove all the spaces. Since we arranged for each result to be in a box, the boxes will align giving us a somewhat readable presentation.
Java
Code:
<lang java>import java.util.*;
class Zeckendorf {
public static String getZeckendorf(int n) { if (n == 0) return "0"; List<Integer> fibNumbers = new ArrayList<Integer>(); fibNumbers.add(1); int nextFib = 2; while (nextFib <= n) { fibNumbers.add(nextFib); nextFib += fibNumbers.get(fibNumbers.size() - 2); } StringBuilder sb = new StringBuilder(); for (int i = fibNumbers.size() - 1; i >= 0; i--) { int fibNumber = fibNumbers.get(i); sb.append((fibNumber <= n) ? "1" : "0"); if (fibNumber <= n) n -= fibNumber; } return sb.toString(); } public static void main(String[] args) { for (int i = 0; i <= 20; i++) System.out.println("Z(" + i + ")=" + getZeckendorf(i)); }
}</lang>
Output:
Z(0)=0 Z(1)=1 Z(2)=10 Z(3)=100 Z(4)=101 Z(5)=1000 Z(6)=1001 Z(7)=1010 Z(8)=10000 Z(9)=10001 Z(10)=10010 Z(11)=10100 Z(12)=10101 Z(13)=100000 Z(14)=100001 Z(15)=100010 Z(16)=100100 Z(17)=100101 Z(18)=101000 Z(19)=101001 Z(20)=101010
Recursive Implementation
Code: <lang java>import java.util.ArrayList; import java.util.List;
public class Zeckendorf {
private List<Integer> getFibList(final int maxNum, final int n1, final int n2, final List<Integer> fibs){ if(n2 > maxNum) return fibs;
fibs.add(n2);
return getFibList(maxNum, n2, n1 + n2, fibs); }
public String getZeckendorf(final int num) { if (num <= 0) return "0";
final List<Integer> fibs = getFibList(num, 1, 2, new ArrayList<Integer>()Template:Add(1););
return getZeckString("", num, fibs.size() - 1, fibs); }
private String getZeckString(final String zeck, final int num, final int index, final List<Integer> fibs){ final int curFib = fibs.get(index); final boolean placeZeck = num >= curFib;
final String outString = placeZeck ? zeck + "1" : zeck + "0"; final int outNum = placeZeck ? num - curFib : num;
if(index == 0) return outString;
return getZeckString(outString, outNum, index - 1, fibs); } public static void main(final String[] args) { final Zeckendorf zeckendorf = new Zeckendorf();
for(int i =0; i <= 20; i++){ System.out.println("Z("+ i +"):\t" + zeckendorf.getZeckendorf(i)); } }
}</lang>
Output:
Z(0): 0 Z(1): 1 Z(2): 10 Z(3): 100 Z(4): 101 Z(5): 1000 Z(6): 1001 Z(7): 1010 Z(8): 10000 Z(9): 10001 Z(10): 10010 Z(11): 10100 Z(12): 10101 Z(13): 100000 Z(14): 100001 Z(15): 100010 Z(16): 100100 Z(17): 100101 Z(18): 101000 Z(19): 101001 Z(20): 101010
JavaScript
ES6
(mapAccumuL example)
<lang JavaScript>(() => {
'use strict';
const main = () => unlines( map(n => concat(zeckendorf(n)), enumFromTo(0, 20) ) );
// zeckendorf :: Int -> String const zeckendorf = n => { const go = (n, x) => n < x ? ( Tuple(n, '0') ) : Tuple(n - x, '1') return 0 < n ? ( snd(mapAccumL( go, n, reverse(fibUntil(n)) )) ) : ['0']; };
// fibUntil :: Int -> [Int] const fibUntil = n => cons(1, takeWhile(x => n >= x, map(snd, iterateUntil( tpl => n <= fst(tpl), tpl => { const x = snd(tpl); return Tuple(x, x + fst(tpl)); }, Tuple(1, 2) ))));
// GENERIC FUNCTIONS ----------------------------
// Tuple (,) :: a -> b -> (a, b) const Tuple = (a, b) => ({ type: 'Tuple', '0': a, '1': b, length: 2 });
// concat :: a -> [a] // concat :: [String] -> String const concat = xs => 0 < xs.length ? (() => { const unit = 'string' !== typeof xs[0] ? ( [] ) : ; return unit.concat.apply(unit, xs); })() : [];
// cons :: a -> [a] -> [a] const cons = (x, xs) => Array.isArray(xs) ? ( [x].concat(xs) ) : (x + xs);
// enumFromTo :: Int -> Int -> [Int] const enumFromTo = (m, n) => m <= n ? iterateUntil( x => n <= x, x => 1 + x, m ) : [];
// fst :: (a, b) -> a const fst = tpl => tpl[0];
// iterateUntil :: (a -> Bool) -> (a -> a) -> a -> [a] const iterateUntil = (p, f, x) => { const vs = [x]; let h = x; while (!p(h))(h = f(h), vs.push(h)); return vs; };
// map :: (a -> b) -> [a] -> [b] const map = (f, xs) => xs.map(f);
// 'The mapAccumL function behaves like a combination of map and foldl; // it applies a function to each element of a list, passing an accumulating // parameter from left to right, and returning a final value of this // accumulator together with the new list.' (See Hoogle)
// mapAccumL :: (acc -> x -> (acc, y)) -> acc -> [x] -> (acc, [y]) const mapAccumL = (f, acc, xs) => xs.reduce((a, x, i) => { const pair = f(a[0], x, i); return Tuple(pair[0], a[1].concat(pair[1])); }, Tuple(acc, []));
// reverse :: [a] -> [a] const reverse = xs => 'string' !== typeof xs ? ( xs.slice(0).reverse() ) : xs.split().reverse().join();
// snd :: (a, b) -> b const snd = tpl => tpl[1];
// tail :: [a] -> [a] const tail = xs => 0 < xs.length ? xs.slice(1) : [];
// takeWhile :: (a -> Bool) -> [a] -> [a] // takeWhile :: (Char -> Bool) -> String -> String const takeWhile = (p, xs) => { const lng = xs.length; return 0 < lng ? xs.slice( 0, until( i => i === lng || !p(xs[i]), i => 1 + i, 0 ) ) : []; };
// unlines :: [String] -> String const unlines = xs => xs.join('\n');
// until :: (a -> Bool) -> (a -> a) -> a -> a const until = (p, f, x) => { let v = x; while (!p(v)) v = f(v); return v; };
// MAIN --- return main();
})();</lang>
- Output:
0 1 10 100 101 1000 1001 1010 10000 10001 10010 10100 10101 100000 100001 100010 100100 100101 101000 101001 101010
jq
<lang jq>def zeckendorf:
# rfibs(n) returns an array of fibonnaci numbers up to n, # beginning with 1, 2, ..., in reverse order def rfibs(n): # input: [f(i-2), f(i-1)] [1,1] | [recurse( if .[1] >= n then empty else [.[1], add] end ) | .[1]] | reverse;
. as $n # [n, rfibs, digit ] | [$n, rfibs($n), "" ] | [ recurse( .[0] as $n | .[1] as $f | if ($f|length) == 0 then empty else $f[0] as $next | if $n >= $next then [ ( $n - $next), $f[1:], "1"]
else [ $n, $f[1:], "0"]
end end ) | .[2] ] | if .[1] == "0" then .[2:] else . end # remove leading 0 if any | join("") ;</lang>
Example: <lang jq>range(0;21) | "\(.): \(zeckendorf)"</lang>
- Output:
<lang sh>$ jq -n -r -f zeckendorf.jq 0: 1: 1 2: 10 3: 100 4: 101 5: 1000 6: 1001 7: 1010 8: 10000 9: 10001 10: 10010 11: 10100 12: 10101 13: 100000 14: 100001 15: 100010 16: 100100 17: 100101 18: 101000 19: 101001 20: 101010</lang>
Julia
<lang julia>function zeck(n)
n <= 0 && return 0 fib = [2,1]; while fib[1] < n unshift!(fib,sum(fib[1:2])) end dig = Int[]; for f in fib f <= n ? (push!(dig,1); n = n-f;) : push!(dig,0) end return dig[1] == 0 ? dig[2:end] : dig
end</lang>
- Output:
julia> for x = 0:20 println(join(zeck(x))) end 0 1 10 100 101 1000 1001 1010 10000 10001 10010 10100 10101 100000 100001 100010 100100 100101 101000 101001 101010
Kotlin
<lang scala>// version 1.0.6
const val LIMIT = 46 // to stay within range of signed 32 bit integer val fibs = fibonacci(LIMIT)
fun fibonacci(n: Int): IntArray {
if (n !in 2..LIMIT) throw IllegalArgumentException("n must be between 2 and $LIMIT") val fibs = IntArray(n) fibs[0] = 1 fibs[1] = 1 for (i in 2 until n) fibs[i] = fibs[i - 1] + fibs[i - 2] return fibs
}
fun zeckendorf(n: Int): String {
if (n < 0) throw IllegalArgumentException("n must be non-negative") if (n < 2) return n.toString() var lastFibIndex = 1 for (i in 2..LIMIT) if (fibs[i] > n) { lastFibIndex = i - 1 break } var nn = n - fibs[lastFibIndex--] val zr = StringBuilder("1") for (i in lastFibIndex downTo 1) if (fibs[i] <= nn) { zr.append('1') nn -= fibs[i] } else { zr.append('0') } return zr.toString()
}
fun main(args: Array<String>) {
println(" n z") for (i in 0..20) println("${"%2d".format(i)} : ${zeckendorf(i)}")
}</lang>
- Output:
n z 0 : 0 1 : 1 2 : 10 3 : 100 4 : 101 5 : 1000 6 : 1001 7 : 1010 8 : 10000 9 : 10001 10 : 10010 11 : 10100 12 : 10101 13 : 100000 14 : 100001 15 : 100010 16 : 100100 17 : 100101 18 : 101000 19 : 101001 20 : 101010
Lingo
<lang Lingo>-- Return the distinct Fibonacci numbers not greater than 'n' on fibsUpTo (n)
fibList = [] last = 1 current = 1 repeat while current <= n fibList.add(current) nxt = last + current last = current current = nxt end repeat return fibList
end
-- Return the Zeckendorf representation of 'n' on zeckendorf (n)
fib = fibsUpTo(n) zeck = "" repeat with pos = fib.count down to 1 if n >= fib[pos] then zeck = zeck & "1" n = n - fib[pos] else zeck = zeck & "0" end if end repeat if zeck = "" then return "0" return zeck
end</lang>
<lang Lingo>repeat with n = 0 to 20
put n & ": " & zeckendorf(n)
end repeat</lang>
- Output:
0: 0 1: 1 2: 10 3: 100 4: 101 5: 1000 6: 1001 7: 1010 8: 10000 9: 10001 10: 10010 11: 10100 12: 10101 13: 100000 14: 100001 15: 100010 16: 100100 17: 100101 18: 101000 19: 101001 20: 101010
Logo
<lang logo>; return the (N+1)th Fibonacci number (1,2,3,5,8,13,...) to fib m
local "n make "n sum :m 1 if [lessequal? :n 0] [output difference fib sum :n 2 fib sum :n 1] global "_fib if [not name? "_fib] [ make "_fib [1 1] ] local "length make "length count :_fib while [greater? :n :length] [ make "_fib (lput (sum (last :_fib) (last (butlast :_fib))) :_fib) make "length sum :length 1 ] output item :n :_fib
end
- return the binary Zeckendorf representation of a nonnegative number
to zeckendorf n
if [less? :n 0] [(throw "error [Number must be nonnegative.])] (local "i "f "result) make "i :n make "f fib :i while [less? :f :n] [make "i sum :i 1 make "f fib :i]
make "result "|| while [greater? :i 0] [ ifelse [greaterequal? :n :f] [ make "result lput 1 :result make "n difference :n :f ] [ if [not empty? :result] [ make "result lput 0 :result ] ] make "i difference :i 1 make "f fib :i ] if [equal? :result "||] [ make "result 0 ] output :result
end
type zeckendorf 0 repeat 20 [
type word "| | zeckendorf repcount
] print [] bye</lang>
- Output:
0 1 10 100 101 1000 1001 1010 10000 10001 10010 10100 10101 100000 100001 100010 100100 100101 101000 101001 101010
Lua
<lang Lua>-- Return the distinct Fibonacci numbers not greater than 'n' function fibsUpTo (n)
local fibList, last, current, nxt = {}, 1, 1 while current <= n do table.insert(fibList, current) nxt = last + current last = current current = nxt end return fibList
end
-- Return the Zeckendorf representation of 'n' function zeckendorf (n)
local fib, zeck = fibsUpTo(n), "" for pos = #fib, 1, -1 do if n >= fib[pos] then zeck = zeck .. "1" n = n - fib[pos] else zeck = zeck .. "0" end end if zeck == "" then return "0" end return zeck
end
-- Main procedure print(" n\t| Zeckendorf(n)") print(string.rep("-", 23)) for n = 0, 20 do
print(" " .. n, "| " .. zeckendorf(n))
end</lang>
- Output:
n | Zeckendorf(n) ----------------------- 0 | 0 1 | 1 2 | 10 3 | 100 4 | 101 5 | 1000 6 | 1001 7 | 1010 8 | 10000 9 | 10001 10 | 10010 11 | 10100 12 | 10101 13 | 100000 14 | 100001 15 | 100010 16 | 100100 17 | 100101 18 | 101000 19 | 101001 20 | 101010
Mathematica
<lang Mathematica>zeckendorf[0] = 0; zeckendorf[n_Integer] :=
10^(# - 1) + zeckendorf[n - Fibonacci[# + 1]] &@ LengthWhile[ Fibonacci /@ Range[2, Ceiling@Log[GoldenRatio, n Sqrt@5]], # <= n &];
zeckendorf /@ Range[0, 20]</lang>
- Output:
{0, 1, 10, 100, 101, 1000, 1001, 1010, 10000, 10001, 10010, 10100, 10101, 100000, 100001, 100010, 100100, 100101, 101000, 101001, 101010}
Nim
<lang nim>import strutils
proc z(n): string =
if n == 0: return "0" var fib = @[2,1] var n = n while fib[0] < n: fib.insert(fib[0] + fib[1]) result = "" for f in fib: if f <= n: result.add '1' n -= f else: result.add '0' if result[0] == '0': result = result[1..result.high]
for i in 0 .. 20:
echo align($i, 3)," ",align(z(i), 8)</lang>
Output:
0 0 1 1 2 10 3 100 4 101 5 1000 6 1001 7 1010 8 10000 9 10001 10 10010 11 10100 12 10101 13 100000 14 100001 15 100010 16 100100 17 100101 18 101000 19 101001 20 101010
PARI/GP
<lang parigp>Z(n)=if(!n,print1(0));my(k=2);while(fibonacci(k)<=n,k++); forstep(i=k-1,2,-1,print1(if(fibonacci(i)<=n,n-=fibonacci(i);1,0)));print for(n=0,20,Z(n))</lang>
0 1 10 100 101 1000 1001 1010 10000 10001 10010 10100 10101 100000 100001 100010 100100 100101 101000 101001 101010
Perl
<lang perl>my @fib;
sub fib {
my $n = shift; return 1 if $n < 2; return $fib[$n] //= fib($n-1)+fib($n-2);
}
sub zeckendorf {
my $n = shift; return "0" unless $n; my $i = 1; $i++ while fib($i) <= $n; my $z = ; while( --$i ) { $z .= "0", next if fib( $i ) > $n; $z .= "1"; $n -= fib( $i ); } return $z;
}
printf "%4d: %8s\n", $_, zeckendorf($_) for 0..20; </lang>
- Output:
0: 0 1: 1 2: 10 3: 100 4: 101 5: 1000 6: 1001 7: 1010 8: 10000 9: 10001 10: 10010 11: 10100 12: 10101 13: 100000 14: 100001 15: 100010 16: 100100 17: 100101 18: 101000 19: 101001 20: 101010
Perl 6
<lang perl6>printf "%2d: %8s\n", $_, zeckendorf($_) for 0 .. 20;
multi zeckendorf(0) { '0' } multi zeckendorf($n is copy) {
constant FIBS = (1,2, *+* ... *).cache; [~] map { $n -= $_ if my $digit = $n >= $_; +$digit; }, reverse FIBS ...^ * > $n;
}</lang>
- Output:
0: 0 1: 1 2: 10 3: 100 4: 101 5: 1000 6: 1001 7: 1010 8: 10000 9: 10001 10: 10010 11: 10100 12: 10101 13: 100000 14: 100001 15: 100010 16: 100100 17: 100101 18: 101000 19: 101001 20: 101010
Phix
<lang Phix>function zeckendorf(integer n) integer r = 0, c sequence fib = {1,1}
while fib[$]<n do fib &= fib[$] + fib[$-1] end while for i=length(fib) to 2 by -1 do c = n>=fib[i] r += r+c n -= c*fib[i] end for return r
end function
for i=0 to 20 do
printf(1,"%2d: %7b\n",{i,zeckendorf(i)})
end for</lang>
- Output:
0: 0 1: 1 2: 10 3: 100 4: 101 5: 1000 6: 1001 7: 1010 8: 10000 9: 10001 10: 10010 11: 10100 12: 10101 13: 100000 14: 100001 15: 100010 16: 100100 17: 100101 18: 101000 19: 101001 20: 101010
PHP
<lang PHP> <?php $m = 20;
$F = array(1,1); while ($F[count($F)-1] <= $m)
$F[] = $F[count($F)-1] + $F[count($F)-2];
while ($n = $m--) {
while ($F[count($F)-1] > $n) array_pop($F); $l = count($F)-1; print "$n: "; while ($n) { if ($n >= $F[$l]) { $n = $n - $F[$l]; print '1'; } else print '0'; --$l; } print str_repeat('0',$l); print "\n";
} ?> </lang>
- Output:
20: 101010 19: 101001 18: 101000 17: 100101 16: 100100 15: 100010 14: 100001 13: 100000 12: 10101 11: 10100 10: 10010 9: 10001 8: 10000 7: 1010 6: 1001 5: 1000 4: 101 3: 100 2: 10 1: 1
PicoLisp
<lang PicoLisp>(de fib (N)
(let Fibs (1 1) (while (>= N (+ (car Fibs) (cadr Fibs))) (push 'Fibs (+ (car Fibs) (cadr Fibs))) ) (uniq Fibs) ) )
(de zecken1 (N)
(make (for I (fib N) (if (> I N) (link 0) (link 1) (dec 'N I) ) ) ) )
(de zecken2 (N)
(make (when (=0 N) (link 0)) (for I (fib N) (when (<= I N) (link I) (dec 'N I) ) ) ) )
(for (N 0 (> 21 N) (inc N))
(tab (2 4 6 2 -10) N " -> " (zecken1 N) " " (glue " + " (zecken2 N)) ) )
(bye)</lang>
- Output:
0 -> 0 0 1 -> 1 1 2 -> 10 2 3 -> 100 3 4 -> 101 3 + 1 5 -> 1000 5 6 -> 1001 5 + 1 7 -> 1010 5 + 2 8 -> 10000 8 9 -> 10001 8 + 1 10 -> 10010 8 + 2 11 -> 10100 8 + 3 12 -> 10101 8 + 3 + 1 13 -> 100000 13 14 -> 100001 13 + 1 15 -> 100010 13 + 2 16 -> 100100 13 + 3 17 -> 100101 13 + 3 + 1 18 -> 101000 13 + 5 19 -> 101001 13 + 5 + 1 20 -> 101010 13 + 5 + 2
Plain TeX
This code needs an etex engine.
<lang tex>\def\genfibolist#1{% #creates the fibo list which sum>=#1 \let\fibolist\empty\def\targetsum{#1}\def\fibosum{0}% \genfibolistaux1,1\relax } \def\genfibolistaux#1,#2\relax{% \ifnum\fibosum<\targetsum\relax \edef\fibosum{\number\numexpr\fibosum+#2}% \edef\fibolist{#2,\fibolist}% \edef\tempfibo{\noexpand\genfibolistaux#2,\number\numexpr#1+#2\relax\relax}% \expandafter\tempfibo \fi } \def\zeckendorf#1{\expandafter\zeckendorfaux\fibolist,\relax#1\relax\relax0} \def\zeckendorfaux#1,#2\relax#3\relax#4\relax#5{% \ifx\relax#2\relax #4% \else \ifnum#3<#1 \edef\temp{#2\relax#3\relax#4\ifnum#5=1 0\fi\relax#5}% \else \edef\temp{#2\relax\number\numexpr#3-#1\relax\relax#41\relax1}% \fi \expandafter\expandafter\expandafter\zeckendorfaux\expandafter\temp \fi } \newcount\ii \def\listzeckendorf#1{% \genfibolist{#1}% \ii=0 \loop \ifnum\ii<#1 \advance\ii1 \number\ii: \zeckendorf\ii\endgraf \repeat } \listzeckendorf{20}% any integer accepted \bye</lang>
pdf output looks like:
1: 1 2: 10 3: 100 4: 101 5: 1000 6: 1001 7: 1010 8: 10000 9: 10001 10: 10010 11: 10100 12: 10101 13: 100000 14: 100001 15: 100010 16: 100100 17: 100101 18: 101000 19: 101001 20: 101010
PowerShell
<lang PowerShell> function Get-ZeckendorfNumber ( $N )
{ # Calculate relevant portation of Fibonacci series $Fib = @( 1, 1 ) While ( $Fib[-1] -lt $N ) { $Fib += $Fib[-1] + $Fib[-2] } # Start with 0 $ZeckendorfNumber = 0 # For each number in the relevant portion of Fibonacci series For ( $i = $Fib.Count - 1; $i -gt 0; $i-- ) { # If Fibonacci number is less than or equal to remainder of N If ( $Fib[$i] -le $N ) { # Double Z number and add 1 (equivalent to adding a '1' to the end of a binary number) $ZeckendorfNumber = $ZeckendorfNumber * 2 + 1 # Reduce N by Fibonacci number, skip next Fibonacci number $N -= $Fib[$i--] } # If were aren't finished yet, double Z number # (equivalent to adding a '0' to the end of a binary number) If ( $i ) { $ZeckendorfNumber *= 2 } } return $ZeckendorfNumber }
</lang> <lang PowerShell>
- Get Zeckendorf numbers through 20, convert to binary for display
0..20 | ForEach { [convert]::ToString( ( Get-ZeckendorfNumber $_ ), 2 ) } </lang>
- Output:
0 1 10 100 101 1000 1001 1010 10000 10001 10010 10100 10101 100000 100001 100010 100100 100101 101000 101001 101010
PureBasic
<lang PureBasic>Procedure.s zeck(n.i)
Dim f.i(1) : Define i.i=1, o$ f(0)=1 : f(1)=1 While f(i)<n i+1 : ReDim f(ArraySize(f())+1) : f(i)=f(i-1)+f(i-2) Wend For i=i To 1 Step -1 If n>=f(i) : o$+"1" : n-f(i) : Else : o$+"0" : EndIf Next If Len(o$)>1 : o$=LTrim(o$,"0") : EndIf ProcedureReturn o$
EndProcedure
Define n.i, t$ OpenConsole("Zeckendorf number representation") PrintN(~"\tNr.\tZeckendorf") For n=0 To 20
t$=zeck(n) If FindString(t$,"11") PrintN("Error: n= "+Str(n)+~"\tZeckendorf= "+t$) Break Else PrintN(~"\t"+RSet(Str(n),3," ")+~"\t"+RSet(t$,7," ")) EndIf
Next Input()</lang>
- Output:
Nr. Zeckendorf 0 0 1 1 2 10 3 100 4 101 5 1000 6 1001 7 1010 8 10000 9 10001 10 10010 11 10100 12 10101 13 100000 14 100001 15 100010 16 100100 17 100101 18 101000 19 101001 20 101010
Python
<lang python>def fib():
memo = [1, 2] while True: memo.append(sum(memo)) yield memo.pop(0)
def sequence_down_from_n(n, seq_generator):
seq = [] for s in seq_generator(): seq.append(s) if s >= n: break return seq[::-1]
def zeckendorf(n):
if n == 0: return [0] seq = sequence_down_from_n(n, fib) digits, nleft = [], n for s in seq: if s <= nleft: digits.append(1) nleft -= s else: digits.append(0) assert nleft == 0, 'Check all of n is accounted for' assert sum(x*y for x,y in zip(digits, seq)) == n, 'Assert digits are correct' while digits[0] == 0: # Remove any zeroes padding L.H.S. digits.pop(0) return digits
n = 20 print('Fibonacci digit multipliers: %r' % sequence_down_from_n(n, fib)) for i in range(n + 1):
print('%3i: %8s' % (i, .join(str(d) for d in zeckendorf(i))))</lang>
- Output:
Fibonacci digit multipliers: [21, 13, 8, 5, 3, 2, 1] 0: 0 1: 1 2: 10 3: 100 4: 101 5: 1000 6: 1001 7: 1010 8: 10000 9: 10001 10: 10010 11: 10100 12: 10101 13: 100000 14: 100001 15: 100010 16: 100100 17: 100101 18: 101000 19: 101001 20: 101010
Shorter version
<lang python>n = 20 def z(n):
if n == 0 : return [0] fib = [2,1] while fib[0] < n: fib[0:0] = [sum(fib[:2])] dig = [] for f in fib: if f <= n: dig, n = dig + [1], n - f else: dig += [0] return dig if dig[0] else dig[1:]
for i in range(n + 1):
print('%3i: %8s' % (i, .join(str(d) for d in z(i))))</lang>
- Output:
0: 0 1: 1 2: 10 3: 100 4: 101 5: 1000 6: 1001 7: 1010 8: 10000 9: 10001 10: 10010 11: 10100 12: 10101 13: 100000 14: 100001 15: 100010 16: 100100 17: 100101 18: 101000 19: 101001 20: 101010
R
<lang R>zeckendorf <- function(number) {
# Get an upper limit on Fibonacci numbers needed to cover number indexOfFibonacciNumber <- function(n) { if (n < 1) { 2 } else { Phi <- (1 + sqrt(5)) / 2 invertClosedFormula <- log(n * sqrt(5)) / log(Phi) ceiling(invertClosedFormula) } }
upperLimit <- indexOfFibonacciNumber(number)
# Return the sequence as digits, sorted descending fibonacciSequenceDigits <- function(n) { fibGenerator <- function(f, ...) { c(f[2], sum(f)) } fibSeq <- Reduce(fibGenerator, 1:n, c(0,1), accumulate=TRUE)
fibNums <- unlist(lapply(fibSeq, head, n=1))
# drop last F0 and F1 and reverse sequence rev(fibNums[-2:-1]) }
digits <- fibonacciSequenceDigits(upperLimit)
isInNumber <- function(digit) { if (number >= digit) { number <<- number - digit 1 } else { 0 } }
zeckSeq <- Map(isInNumber, digits)
# drop leading 0 and convert to String gsub("^0+1", "1", paste(zeckSeq, collapse=""))
}
print(unlist(lapply(0:20, zeckendorf)))</lang>
This is definitely not the shortest way to implement the Zeckendorf numbers but focus was on the functional aspect of R, so no loops and (almost) no assignments.
- Output:
[1] "0" "1" "10" "100" "101" "1000" "1001" "1010" [9] "10000" "10001" "10010" "10100" "10101" "100000" "100001" "100010" [17] "100100" "100101" "101000" "101001" "101010"
Racket
<lang racket>
- lang racket (require math)
(define (fibs n)
(reverse (for/list ([i (in-naturals 2)] #:break (> (fibonacci i) n)) (fibonacci i))))
(define (zechendorf n)
(match/values (for/fold ([n n] [xs '()]) ([f (fibs n)]) (if (> f n) (values n (cons 0 xs)) (values (- n f) (cons 1 xs)))) [(_ xs) (reverse xs)]))
(for/list ([n 21])
(list n (zechendorf n)))
</lang> Output: <lang racket> '((0 ())
(1 (1)) (2 (1 0)) (3 (1 0 0)) (4 (1 0 1)) (5 (1 0 0 0)) (6 (1 0 0 1)) (7 (1 0 1 0)) (8 (1 0 0 0 0)) (9 (1 0 0 0 1)) (10 (1 0 0 1 0)) (11 (1 0 1 0 0)) (12 (1 0 1 0 1)) (13 (1 0 0 0 0 0)) (14 (1 0 0 0 0 1)) (15 (1 0 0 0 1 0)) (16 (1 0 0 1 0 0)) (17 (1 0 0 1 0 1)) (18 (1 0 1 0 0 0)) (19 (1 0 1 0 0 1)) (20 (1 0 1 0 1 0)))
</lang>
REXX
specific to 20
<lang rexx> /* REXX ***************************************************************
- 11.10.2012 Walter Pachl
- /
fib='13 8 5 3 2 1' Do i=6 To 1 By -1 /* Prepare Fibonacci Numbers */
Parse Var fib f.i fib /* f.1 ... f.7 */ End
Do n=0 To 20 /* for all numbers in the task */
m=n /* copy of number */ r= /* result for n */ Do i=6 To 1 By -1 /* loop through numbers */ If m>=f.i Then Do /* f.i must be used */ r=r||1 /* 1 into result */ m=m-f.i /* subtract */ End Else /* f.i is larger than the rest */ r=r||0 /* 0 into result */ End r=strip(r,'L','0') /* strip leading zeros */ If r= Then r='0' /* take care of 0 */ Say right(n,2)': 'right(r,6) /* show result */ End</lang>
Output:
0: 0 1: 1 2: 10 3: 100 4: 101 5: 1000 6: 1001 7: 1010 8: 10000 9: 10001 10: 10010 11: 10100 12: 10101 13: 100000 14: 100001 15: 100010 16: 100100 17: 100101 18: 101000 19: 101001 20: 101010
generalized
This generalized REXX version will work for any Zeckendorf number (up to 100,000 decimal digits).
A list of Fibonacci numbers (in ascending order) is generated large enough to handle the Nth Zeckendorf number.
<lang rexx>/*REXX program calculates and displays the first N Zeckendorf numbers. */
numeric digits 100000 /*just in case user gets real ka─razy. */
parse arg N . /*let the user specify the upper limit.*/
if N== | N=="," then n=20; w= length(N) /*Not specified? Then use the default.*/
@.1= 1 /*start the array with 1 and 2. */
@.2= 2; do #=3 until #>=N; p= #-1; pp= #-2 /*build a list of Fibonacci numbers. */
@.#= @.p + @.pp /*sum the last two Fibonacci numbers. */ end /*#*/ /* [↑] #: contains a Fibonacci list.*/
do j=0 to N; parse var j x z /*task: process zero ──► N numbers.*/ do k=# by -1 for #; _= @.k /*process all the Fibonacci numbers. */ if x>=_ then do; z= z'1' /*is X>the next Fibonacci #? Append 1.*/ x= x - _ /*subtract this Fibonacci # from index.*/ end else z= z'0' /*append zero (0) to the Fibonacci #. */ end /*k*/ say ' Zeckendorf' right(j, w) "=" right(z+0, 30) /*display a number.*/ end /*j*/ /*stick a fork in it, we're all done. */</lang>
- output when using the default input:
Zeckendorf 0 = 0 Zeckendorf 1 = 1 Zeckendorf 2 = 10 Zeckendorf 3 = 100 Zeckendorf 4 = 101 Zeckendorf 5 = 1000 Zeckendorf 6 = 1001 Zeckendorf 7 = 1010 Zeckendorf 8 = 10000 Zeckendorf 9 = 10001 Zeckendorf 10 = 10010 Zeckendorf 11 = 10100 Zeckendorf 12 = 10101 Zeckendorf 13 = 100000 Zeckendorf 14 = 100001 Zeckendorf 15 = 100010 Zeckendorf 16 = 100100 Zeckendorf 17 = 100101 Zeckendorf 18 = 101000 Zeckendorf 19 = 101001 Zeckendorf 20 = 101010
generic
This generic REXX version will generate up to the Nth Zeckendorf numbers (up to 100,000 decimal digits) by
using binary numbers that don't have two consecutive 11s within their binary version.
There isn't any need to generate a Fibonacci series with this method. This method is extremely fast. <lang REXX>/*REXX program calculates and displays the first N Zeckendorf numbers. */ numeric digits 100000 /*just in case user gets real ka─razy. */ parse arg N . /*let the user specify the upper limit.*/ if N== | N=="," then n=20; w= length(N) /*Not specified? Then use the default.*/ z=0 /*the index of a Zeckendorf number. */
do j=0 until z>N; _=x2b( d2x(j) ) /*task: process zero ──► N. */ if pos(11, _) \== 0 then iterate /*are there two consecutive ones (1s) ?*/ say ' Zeckendorf' right(z, w) "=" right(_+0, 30) /*display a number.*/ z= z + 1 /*bump the Zeckendorf number counter.*/ end /*j*/ /*stick a fork in it, we're all done. */</lang>
{{out|output|text= is identical to the previous (generalized) version.
Ring
<lang ring>
- Project : Zeckendorf number representation
see "0 0" + nl for n = 1 to 20
see "" + n + " " + zeckendorf(n) + nl
next
func zeckendorf(n)
fib = list(45) fib[1] = 1 fib[2] = 1 i = 2 o = "" while fib[i] <= n i = i + 1 fib[i] = fib[i-1] + fib[i-2] end while i != 2 i = i - 1 if n >= fib[i] o = o + "1" n = n - fib[i] else o = o + "0" ok end return o
</lang> Output:
0 0 1 1 2 10 3 100 4 101 5 1000 6 1001 7 1010 8 10000 9 10001 10 10010 11 10100 12 10101 13 100000 14 100001 15 100010 16 100100 17 100101 18 101000 19 101001 20 101010
Ruby
Featuring a method doubling as an enumerator. <lang ruby>def zeckendorf
return to_enum(__method__) unless block_given? x = 0 loop do bin = x.to_s(2) yield bin unless bin.include?("11") x += 1 end
end
zeckendorf.take(21).each_with_index{|x,i| puts "%3d: %8s"% [i, x]} </lang>
- Output:
0: 0 1: 1 2: 10 3: 100 4: 101 5: 1000 6: 1001 7: 1010 8: 10000 9: 10001 10: 10010 11: 10100 12: 10101 13: 100000 14: 100001 15: 100010 16: 100100 17: 100101 18: 101000 19: 101001 20: 101010
<lang ruby>def zeckendorf(n)
return 0 if n.zero? fib = [1,2] fib << fib[-2] + fib[-1] while fib[-1] < n dig = "" fib.reverse_each do |f| if f <= n dig, n = dig + "1", n - f else dig += "0" end end dig.to_i
end
for i in 0..20
puts '%3d: %8d' % [i, zeckendorf(i)]
end</lang> (Same output.)
Scala
<lang scala>def zNum( n:BigInt ) : String = {
if( n == 0 ) return "0" // Short-circuit this and return zero if we were given zero
val v = n.abs
val fibs : Stream[BigInt] = { def series(i:BigInt,j:BigInt):Stream[BigInt] = i #:: series(j, i+j); series(1,0).tail.tail.tail }
def z( v:BigInt ) : List[BigInt] = if(v == 0) List() else {val m = fibs(fibs.indexWhere(_>v) - 1); m :: z(v-m)}
val zv = z(v) // Walk the list of fibonacci numbers from the number that matches the most significant down to 1, // if the zeckendorf matchs then yield '1' otherwise '0' val s = (for( i <- (fibs.indexWhere(_==zv(0)) to 0 by -1) ) yield { if( zv.contains(fibs(i))) "1" else "0"
}).mkString if( n < 0 ) "-" + s // Using a negative-sign instead of twos-complement else s
}
// A little test...
(0 to 20) foreach( i => print( zNum(i) + "\n" ) )
</lang>
- Output:
0 1 10 100 101 1000 1001 1010 10000 10001 10010 10100 10101 100000 100001 100010 100100 100101 101000 101001 101010
Scheme
<lang scheme>(import (rnrs))
(define (getFibList maxNum n1 n2 fibs)
(if (> n2 maxNum) fibs (getFibList maxNum n2 (+ n1 n2) (cons n2 fibs))))
(define (getZeckendorf num)
(if (<= num 0) "0" (let ((fibs (getFibList num 1 2 (list 1)))) (getZeckString "" num fibs))))
(define (getZeckString zeck num fibs)
(let* ((curFib (car fibs)) (placeZeck (>= num curFib)) (outString (string-append zeck (if placeZeck "1" "0"))) (outNum (if placeZeck (- num curFib) num))) (if (null? (cdr fibs)) outString (getZeckString outString outNum (cdr fibs)))))
(let loop ((i 0))
(when (<= i 20) (for-each (lambda (item) (display item)) (list "Z(" i "):\t" (getZeckendorf i))) (newline) (loop (+ i 1))))
</lang>
- Output:
Z(0): 0 Z(1): 1 Z(2): 10 Z(3): 100 Z(4): 101 Z(5): 1000 Z(6): 1001 Z(7): 1010 Z(8): 10000 Z(9): 10001 Z(10): 10010 Z(11): 10100 Z(12): 10101 Z(13): 100000 Z(14): 100001 Z(15): 100010 Z(16): 100100 Z(17): 100101 Z(18): 101000 Z(19): 101001 Z(20): 101010
Sidef
<lang ruby>func fib(n) is cached {
n < 2 ? 1 : (fib(n-1) + fib(n-2))
}
func zeckendorf(n) {
n == 0 && return '0' var i = 1 ++i while (fib(i) <= n) gather { while (--i > 0) { var f = fib(i) f > n ? (take '0') : (take '1'; n -= f) } }.join
}
for n (0..20) {
printf("%4d: %8s\n", n, zeckendorf(n))
}</lang>
- Output:
0: 0 1: 1 2: 10 3: 100 4: 101 5: 1000 6: 1001 7: 1010 8: 10000 9: 10001 10: 10010 11: 10100 12: 10101 13: 100000 14: 100001 15: 100010 16: 100100 17: 100101 18: 101000 19: 101001 20: 101010
Simula
<lang simula>BEGIN
INTEGER N, F0, F1, F2, D; N := 20; COMMENT CALCULATE D FROM ANY GIVEN N ; F1 := 1; F2 := 2; F0 := F1 + F2; D := 2; WHILE F0 < N DO BEGIN F1 := F2; F2 := F0; F0 := F1 + F2; D := D + 1; END; BEGIN COMMENT Sinclair ZX81 BASIC Solution ; TEXT Z1, S1; INTEGER I, J, Z; INTEGER ARRAY F(1:D); ! 10 dim f(6) ; F(1) := 1; ! 20 let f(1)=1 ; F(2) := 2; ! 30 let f(2)=2 ; FOR I := 3 STEP 1 UNTIL D DO BEGIN ! 40 for i=3 to 6 ; F(I) := F(I-2) + F(I-1); ! 50 let f(i)=f(i-2)+f(i-1) ; END; ! 60 next i ; FOR I := 0 STEP 1 UNTIL N DO BEGIN ! 70 for i=0 to 20 ; Z1 :- ""; ! 80 let z$="" ; S1 :- " "; ! 90 let s$=" " ; Z := I; ! 100 let z=i ; FOR J := D STEP -1 UNTIL 1 DO BEGIN ! 110 for j=6 to 1 step -1 ; IF J=1 THEN S1 :- "0"; ! 120 if j=1 then let s$="0" ; IF NOT (Z<F(J)) THEN BEGIN ! 130 if z<f(j) then goto 180 ; Z1 :- Z1 & "1"; ! 140 let z$=z$+"1" ; Z := Z-F(J); ! 150 let z=z-f(j) ; S1 :- "0"; ! 160 let s$="0" ; END ELSE ! 170 goto 190 ; Z1 :- Z1 & S1; ! 180 let z$=z$+s$ ; END; ! 190 next j ; OUTINT(I, 0); OUTCHAR(' '); ! 200 print i ; !" "; !; IF I<10 THEN OUTCHAR(' '); ! 210 if i<10 then print " "; !; OUTTEXT(Z1); OUTIMAGE; ! 220 print z$ ; END; ! 230 next i ; END;
END</lang>
- Output:
0 0 1 1 2 10 3 100 4 101 5 1000 6 1001 7 1010 8 10000 9 10001 10 10010 11 10100 12 10101 13 100000 14 100001 15 100010 16 100100 17 100101 18 101000 19 101001 20 101010
Sinclair ZX81 BASIC
Works on the 1k RAM model, albeit without much room for manoeuvre. (You'd like the Zeckendorf numbers further over towards the right-hand side of the screen? Sorry, can't spare the video RAM.) If you have 2k or more, you can replace the constant 6 with some higher value wherever it occurs in the program and enable yourself to represent bigger numbers in Zeckendorf form. <lang basic> 10 DIM F(6)
20 LET F(1)=1 30 LET F(2)=2 40 FOR I=3 TO 6 50 LET F(I)=F(I-2)+F(I-1) 60 NEXT I 70 FOR I=0 TO 20 80 LET Z$="" 90 LET S$=" "
100 LET Z=I 110 FOR J=6 TO 1 STEP -1 120 IF J=1 THEN LET S$="0" 130 IF Z<F(J) THEN GOTO 180 140 LET Z$=Z$+"1" 150 LET Z=Z-F(J) 160 LET S$="0" 170 GOTO 190 180 LET Z$=Z$+S$ 190 NEXT J 200 PRINT I;" "; 210 IF I<10 THEN PRINT " "; 220 PRINT Z$ 230 NEXT I</lang>
- Output:
0 0 1 1 2 10 3 100 4 101 5 1000 6 1001 7 1010 8 10000 9 10001 10 10010 11 10100 12 10101 13 100000 14 100001 15 100010 16 100100 17 100101 18 101000 19 101001 20 101010
Tcl
<lang tcl>package require Tcl 8.5
- Generates the Fibonacci sequence (starting at 1) up to the largest item that
- is no larger than the target value. Could use tricks to precompute, but this
- is actually a pretty cheap linear operation.
proc fibseq target {
set seq {}; set prev 1; set fib 1 for {set n 1;set i 1} {$fib <= $target} {incr n} {
for {} {$i < $n} {incr i} { lassign [list $fib [incr fib $prev]] prev fib } if {$fib <= $target} { lappend seq $fib }
} return $seq
}
- Produce the given Zeckendorf number.
proc zeckendorf n {
# Special case: only value that begins with 0 if {$n == 0} {return 0} set zs {} foreach f [lreverse [fibseq $n]] {
lappend zs [set z [expr {$f <= $n}]] if {$z} {incr n [expr {-$f}]}
} return [join $zs ""]
}</lang> Demonstration <lang tcl>for {set i 0} {$i <= 20} {incr i} {
puts [format "%2d:%9s" $i [zeckendorf $i]]
}</lang>
- Output:
0: 0 1: 1 2: 10 3: 100 4: 101 5: 1000 6: 1001 7: 1010 8: 10000 9: 10001 10: 10010 11: 10100 12: 10101 13: 100000 14: 100001 15: 100010 16: 100100 17: 100101 18: 101000 19: 101001 20: 101010
uBasic/4tH
<lang>For x = 0 to 20 ' Print Zeckendorf numbers 0 - 20
Print x, Push x : Gosub _Zeckendorf ' get Zeckendorf number repres. Print ' terminate line
Next
End
_Fibonacci
Push Tos() ' duplicate TOS() @(0) = 0 ' This function returns the @(1) = 1 ' Fibonacci number which is smaller ' or equal to TOS() Do While @(1) < Tos() + 1 Push (@(1)) @(1) = @(0) + @(1) ' get next Fibonacci number @(0) = Pop() Loop ' loop if not exceeded TOS()
Gosub _Drop ' clear TOS() Push @(0) ' return Fibonacci number
Return
_Zeckendorf
GoSub _Fibonacci ' This function breaks TOS() up Print Tos(); ' into its Zeckendorf components Push -(Pop() - Pop()) ' first digit is always there ' the remainder to resolve Do While Tos() ' now go for the next digits GoSub _Fibonacci Print " + ";Tos(); ' print the next digit Push -(Pop() - Pop()) Loop
Gosub _Drop ' clear TOS()
Return ' and return
_Drop
If Pop()%1 = 0 Then Return ' This function clears TOS()</lang>
Output:
0 0 1 1 2 2 3 3 4 3 + 1 5 5 6 5 + 1 7 5 + 2 8 8 9 8 + 1 10 8 + 2 11 8 + 3 12 8 + 3 + 1 13 13 14 13 + 1 15 13 + 2 16 13 + 3 17 13 + 3 + 1 18 13 + 5 19 13 + 5 + 1 20 13 + 5 + 2 0 OK, 0:901
VBScript
<lang vb> Function Zeckendorf(n) num = n Set fibonacci = CreateObject("System.Collections.Arraylist") fibonacci.Add 1 : fibonacci.Add 2 i = 1 Do While fibonacci(i) < num fibonacci.Add fibonacci(i) + fibonacci(i-1) i = i + 1 Loop tmp = "" For j = fibonacci.Count-1 To 0 Step -1 If fibonacci(j) <= num And (tmp = "" Or Left(tmp,1) <> "1") Then tmp = tmp & "1" num = num - fibonacci(j) Else tmp = tmp & "0" End If Next Zeckendorf = CLng(tmp) End Function
'testing the function For k = 0 To 20 WScript.StdOut.WriteLine k & ": " & Zeckendorf(k) Next </lang>
- Output:
0: 0 1: 1 2: 10 3: 100 4: 101 5: 1000 6: 1001 7: 1010 8: 10000 9: 10001 10: 10010 11: 10100 12: 10101 13: 100000 14: 100001 15: 100010 16: 100100 17: 100101 18: 101000 19: 101001 20: 101010
XPL0
<lang XPL0>include c:\cxpl\codes; \intrinsic 'code' declarations
proc Zeckendorf(N); \Display Zeckendorf number (N <= 20) int N; int Fib, LZ, I; [Fib:= [1, 2, 3, 5, 8, 13]; \Fibonacci sequence LZ:= true; \suppress leading zeros for I:= 5 downto 1 do
[if N >= Fib(I) then [N:= N-Fib(I); ChOut(0, ^1); LZ:= false] else ChOut(0, if LZ then ^ else ^0); ];
ChOut(0, N+^0); \output final digit, which can be 0 ];
int N; [for N:= 0 to 20 do
[if N<10 then ChOut(0,^ ); IntOut(0, N); Text(0, ": "); Zeckendorf(N); CrLf(0); ];
]</lang>
Output:
0: 0 1: 1 2: 10 3: 100 4: 101 5: 1000 6: 1001 7: 1010 8: 10000 9: 10001 10: 10010 11: 10100 12: 10101 13: 100000 14: 100001 15: 100010 16: 100100 17: 100101 18: 101000 19: 101001 20: 101010
Yabasic
<lang Yabasic>sub Zeckendorf(n) local i, n$, c
do n$ = bin$(i) if not instr(n$,"11") then print c,":\t",n$ if c = n break c = c + 1 end if i = i + 1 loop end sub
Zeckendorf(20) </lang>
zkl
<lang zkl> // return powers (0|1) of fib sequence (1,2,3,5,8...) that sum to n fcn zeckendorf(n){ //-->String of 1s & 0s, no consecutive 1's
if(n<=0) return("0"); fibs:=fcn(ab){ ab.append(ab.sum()).pop(0) }.fp(L(1,2)); (0).pump(*,List,fibs,'wrap(fib){ if(fib>n)Void.Stop else fib }) .reverse() .pump(String,fcn(fib,rn){ if(fib>rn.value)"0" else { rn.set(rn.value-fib); "1" } }.fp1(Ref(n)))
}</lang> <lang zkl>[0..20].pump(Console.println,fcn(n){ "%2d: %8s".fmt(n,zeckendorf(n)) });</lang>
- Output:
0: 0 1: 1 2: 10 3: 100 4: 101 5: 1000 6: 1001 7: 1010 8: 10000 9: 10001 10: 10010 11: 10100 12: 10101 13: 100000 14: 100001 15: 100010 16: 100100 17: 100101 18: 101000 19: 101001 20: 101010
- Programming Tasks
- Solutions by Programming Task
- 360 Assembly
- Ada
- ALGOL 68
- AppleScript
- AutoHotkey
- AutoIt
- BBC BASIC
- Befunge
- C
- C++
- C sharp
- Clojure
- Common Lisp
- D
- EchoLisp
- Elena
- Elixir
- F Sharp
- Factor
- Forth
- Fortran
- FreeBASIC
- Go
- Haskell
- J
- Java
- JavaScript
- Jq
- Julia
- Kotlin
- Lingo
- Logo
- Lua
- Mathematica
- Nim
- PARI/GP
- Perl
- Perl 6
- Phix
- PHP
- PicoLisp
- PlainTeX
- PowerShell
- PureBasic
- Python
- R
- Racket
- REXX
- Ring
- Ruby
- Scala
- Scheme
- Sidef
- Simula
- Sinclair ZX81 BASIC
- Tcl
- UBasic/4tH
- VBScript
- XPL0
- Yabasic
- Zkl
- Bitwise operations