Van Eck sequence: Difference between revisions
MaiconSoft (talk | contribs) m Added Delphi reference to Pascal code |
Added Algol 68 |
||
Line 402: | Line 402: | ||
<pre>Element 1 .. 10 of Van Eck sequence: 0 0 1 0 2 0 2 2 1 6 |
<pre>Element 1 .. 10 of Van Eck sequence: 0 0 1 0 2 0 2 2 1 6 |
||
Element 991 .. 1000 of Van Eck sequence: 4 7 30 25 67 225 488 0 10 136</pre> |
Element 991 .. 1000 of Van Eck sequence: 4 7 30 25 67 225 488 0 10 136</pre> |
||
=={{header|ALGOL 68}}== |
|||
<lang algol68>BEGIN # find elements of the Van Eck Sequence - first term is 0, following # |
|||
# terms are 0 if the previous was the first appearance of the element # |
|||
# or how far back in the sequence the last element appeared # |
|||
# returns the first n elements of the Van Eck sequence # |
|||
OP VANECK = ( INT n )[]INT: |
|||
BEGIN |
|||
[ 1 : IF n < 0 THEN 0 ELSE n FI ]INT result; FOR i TO n DO result[ i ] := 0 OD; |
|||
[ 0 : UPB result ]INT pos; FOR i FROM 0 TO n DO pos[ i ] := 0 OD; |
|||
FOR i FROM 2 TO n DO |
|||
INT j = i - 1; |
|||
INT prev = result[ j ]; |
|||
IF pos[ prev ] /= 0 THEN |
|||
# not a new element # |
|||
result[ i ] := j - pos[ prev ] |
|||
FI; |
|||
pos[ prev ] := j |
|||
OD; |
|||
result |
|||
END # VANECK # ; |
|||
# construct the first 1000 terms of the sequence # |
|||
[]INT seq = VANECK 1000; |
|||
# show the first and last 10 elements # |
|||
FOR i TO 10 DO print( ( " ", whole( seq[ i ], 0 ) ) ) OD; |
|||
print( ( newline ) ); |
|||
FOR i FROM UPB seq - 9 TO UPB seq DO print( ( " ", whole( seq[ i ], 0 ) ) ) OD; |
|||
print( ( newline ) ) |
|||
END</lang> |
|||
{{out}} |
|||
<pre> |
|||
0 0 1 0 2 0 2 2 1 6 |
|||
4 7 30 25 67 225 488 0 10 136 |
|||
</pre> |
|||
=={{header|ALGOL-M}}== |
=={{header|ALGOL-M}}== |
Revision as of 20:34, 16 June 2021
![Task](http://static.miraheze.org/rosettacodewiki/thumb/b/ba/Rcode-button-task-crushed.png/64px-Rcode-button-task-crushed.png)
You are encouraged to solve this task according to the task description, using any language you may know.
The sequence is generated by following this pseudo-code:
A: The first term is zero. Repeatedly apply: If the last term is *new* to the sequence so far then: B: The next term is zero. Otherwise: C: The next term is how far back this last term occured previousely.
- Example
Using A:
0
Using B:
0 0
Using C:
0 0 1
Using B:
0 0 1 0
Using C: (zero last occured two steps back - before the one)
0 0 1 0 2
Using B:
0 0 1 0 2 0
Using C: (two last occured two steps back - before the zero)
0 0 1 0 2 0 2 2
Using C: (two last occured one step back)
0 0 1 0 2 0 2 2 1
Using C: (one last appeared six steps back)
0 0 1 0 2 0 2 2 1 6
...
- Task
- Create a function/proceedure/method/subroutine/... to generate the Van Eck sequence of numbers.
- Use it to display here, on this page:
- The first ten terms of the sequence.
- Terms 991 - to - 1000 of the sequence.
- References
- Don't Know (the Van Eck Sequence) - Numberphile video.
- Wikipedia Article: Van Eck's Sequence.
- OEIS sequence: A181391.
11l
<lang 11l>F van_eck(c)
[Int] r V n = 0 V seen = [0] V val = 0 L r.append(val) I r.len == c R r I val C seen[1..] val = seen.index(val, 1) E val = 0 seen.insert(0, val) n++
print(‘Van Eck: first 10 terms: ’van_eck(10)) print(‘Van Eck: terms 991 - 1000: ’van_eck(1000)[(len)-10..])</lang>
- Output:
Van Eck: first 10 terms: [0, 0, 1, 0, 2, 0, 2, 2, 1, 6] Van Eck: terms 991 - 1000: [4, 7, 30, 25, 67, 225, 488, 0, 10, 136]
8080 Assembly
<lang 8080asm> org 100h lxi h,ecks ; Zero out 2000 bytes lxi b,0 lxi d,2000 zero: mov m,b inx h dcx d mov a,d ora e jnz zero lxi b,-1 ; BC = Outer loop variable outer: inx b mvi a,3 ; Are we there yet? 1000 = 03E8h cmp b ; Compare high byte jnz go mvi a,0E8h ; Compare low byte cmp c jz done go: mov d,b ; DE = Inner loop variable mov e,c inner: dcx d mov a,d ; <= 0? ral jc outer push b ; Keep both pointers push d mov h,b ; Load BC = eck[BC] mov l,c call eck mov c,m inx h mov b,m xchg ; Load HL = -eck[DE] call eck xchg ldax d cma mov l,a inx d ldax d cma mov h,a inx h ; Two's complement dad b ; -eck[DE] + eck[BC] mov a,h ; Unfortunately this does not set flags ora l ; Check zero pop d ; Meanwhile, restore the pointers pop b jnz inner ; If no match, continue with inner loop mov h,b ; If we _did_, then get &eck[BC + 1] mov l,c inx h call eck mov a,c ; Store BC - DE at that address sub e mov m,a inx h mov a,b sbb d mov m,a jmp outer ; And continue the outer loop done: lxi h,0 ; Print first 10 terms call p10 lxi h,990 ; Print last 10 terms p10: mvi b,10 ; Print 10 terms starting at term HL call eck ploop: mov e,m ; Load term into DE inx h mov d,m inx h push b ; Keep counter push h ; Keep pointer xchg ; Term in HL call printn ; Print term pop h ; Restore pointer and counter pop b dcr b jnz ploop lxi d,nl ; Print a newline afterwards jmp prints eck: push b ; Set HL = &eck[HL] lxi b,ecks ; Base address dad h ; Multiply by two dad b ; Add base pop b ret printn: lxi d,buf ; Print the number in HL push d ; Buffer pointer on stack lxi b,-10 ; Divisor pdigit: lxi d,-1 ; Quotient pdiv: inx d dad b jc pdiv mvi a,'0'+10 add l ; Make ASCII digit pop h dcx h ; Store digit mov m,a push h xchg mov a,h ; Quotient nonzero? ora l jnz pdigit ; Then there are more digits pop d ; Otherwise, print string using CP/M prints: mvi c,9 jmp 5 nl: db 13,10,'$' db '.....' buf: db ' $' ecks: equ $</lang>
- Output:
0 0 1 0 2 0 2 2 1 6 4 7 30 25 67 225 488 0 10 136
8086 Assembly
<lang asm>LIMIT: equ 1000 cpu 8086 org 100h section .text mov di,eck ; Zero out the memory xor ax,ax mov cx,LIMIT rep stosw mov bx,eck ; Base address mov cx,LIMIT ; Limit xor ax,ax mov si,-1 ; Outer loop index outer: inc si dec cx jcxz done mov di,si ; Inner loop index inner: dec di js outer shl si,1 ; Shift the loop indices (each entry is 2 bytes) shl di,1 mov ax,[si+bx] ; Find a match? cmp ax,[di+bx] je match shr si,1 ; If not, shift SI and DI back and keep going shr di,1 jmp inner match: mov ax,si ; Calculate the new value sub ax,di shr ax,1 ; Compensate for shift mov [si+bx+2],ax ; Store value shr si,1 ; Shift SI back and calculate next value jmp outer done: xor si,si ; Print first 10 elements call p10 mov si,LIMIT-10 ; Print last 10 elements⌈ p10: mov cx,10 ; Print 10 elements starting at SI shl si,1 ; Items are 2 bytes wide add si,eck .item: lodsw ; Retrieve item call printn ; Print it loop .item mov dx,nl ; Print a newline afterwards jmp prints printn: mov bx,buf ; Print AX mov bp,10 .digit: xor dx,dx ; Extract digit div bp add dl,'0' ; ASCII digit dec bx mov [bx],dl ; Store in buffer test ax,ax ; Any more digits? jnz .digit mov dx,bx prints: mov ah,9 ; Print string in buffer int 21h ret section .data nl: db 13,10,'$' db '.....' buf: db ' $' section .bss eck: resw LIMIT</lang>
- Output:
0 0 1 0 2 0 2 2 1 6 4 7 30 25 67 225 488 0 10 136
AArch64 Assembly
<lang AArch64 Assembly> /* ARM assembly AARCH64 Raspberry PI 3B */ /* program vanEckSerie64.s */
/*******************************************/ /* Constantes file */ /*******************************************/ /* for this file see task include a file in language AArch64 assembly */ .include "../includeConstantesARM64.inc"
.equ MAXI, 1000
/*********************************/ /* Initialized data */ /*********************************/ .data sMessResultElement: .asciz " @ " szCarriageReturn: .asciz "\n"
/*********************************/ /* UnInitialized data */ /*********************************/ .bss sZoneConv: .skip 24 TableVanEck: .skip 8 * MAXI /*********************************/ /* code section */ /*********************************/ .text .global main main: // entry of program
mov x2,#0 // begin first element mov x3,#0 // current counter ldr x4,qAdrTableVanEck // table address str x2,[x4,x3,lsl 3] // store first zéro
1: // begin loop
mov x5,x3 // init current indice
2:
sub x5,x5,1 // decrement cmp x5,0 // end table ? blt 3f ldr x6,[x4,x5,lsl 3] // load element cmp x6,x2 // and compare with the last element bne 2b // not equal sub x2,x3,x5 // else compute gap b 4f
3:
mov x2,#0 // first, move zero to next element
4:
add x3,x3,#1 // increment counter str x2,[x4,x3,lsl 3] // and store new element cmp x3,MAXI blt 1b mov x2,0
5: // loop display ten elements
ldr x0,[x4,x2,lsl 3] ldr x1,qAdrsZoneConv bl conversion10 // call décimal conversion ldr x0,qAdrsMessResultElement ldr x1,qAdrsZoneConv // insert conversion in message bl strInsertAtCharInc mov x1,0 // final zéro strb w1,[x0,5] // bl affichageMess // display message add x2,x2,1 // increment indice cmp x2,10 // end ? blt 5b // no -> loop ldr x0,qAdrszCarriageReturn bl affichageMess mov x2,MAXI - 10
6: // loop display ten elements 990-999
ldr x0,[x4,x2,lsl 3] ldr x1,qAdrsZoneConv bl conversion10 // call décimal conversion ldr x0,qAdrsMessResultElement ldr x1,qAdrsZoneConv // insert conversion in message bl strInsertAtCharInc mov x1,0 // final zéro strb w1,[x0,5] // bl affichageMess // display message add x2,x2,1 // increment indice cmp x2,MAXI // end ? blt 6b // no -> loop ldr x0,qAdrszCarriageReturn bl affichageMess
100: // standard end of the program
mov x0, 0 // return code mov x8, EXIT // request to exit program svc 0 // perform the system call
qAdrszCarriageReturn: .quad szCarriageReturn qAdrsMessResultElement: .quad sMessResultElement qAdrsZoneConv: .quad sZoneConv qAdrTableVanEck: .quad TableVanEck /********************************************************/ /* File Include fonctions */ /********************************************************/ /* for this file see task include a file in language AArch64 assembly */ .include "../includeARM64.inc" </lang>
0 0 1 0 2 0 2 2 1 6 4 7 30 25 67 225 488 0 10 136
Ada
<lang Ada>with Ada.Text_IO;
procedure Van_Eck_Sequence is
Sequence : array (Natural range 1 .. 1_000) of Natural;
procedure Calculate_Sequence is begin Sequence (Sequence'First) := 0; for Index in Sequence'First .. Sequence'Last - 1 loop Sequence (Index + 1) := 0; for I in reverse Sequence'First .. Index - 1 loop if Sequence (I) = Sequence (Index) then Sequence (Index + 1) := Index - I; exit; end if; end loop; end loop; end Calculate_Sequence;
procedure Show (First, Last : in Positive) is use Ada.Text_IO; begin Put ("Element" & First'Image & " .." & Last'Image & " of Van Eck sequence: "); for I in First .. Last loop Put (Sequence (I)'Image); end loop; New_Line; end Show;
begin
Calculate_Sequence; Show (First => 1, Last => 10); Show (First => 991, Last => 1_000);
end Van_Eck_Sequence;</lang>
- Output:
Element 1 .. 10 of Van Eck sequence: 0 0 1 0 2 0 2 2 1 6 Element 991 .. 1000 of Van Eck sequence: 4 7 30 25 67 225 488 0 10 136
ALGOL 68
<lang algol68>BEGIN # find elements of the Van Eck Sequence - first term is 0, following #
# terms are 0 if the previous was the first appearance of the element # # or how far back in the sequence the last element appeared # # returns the first n elements of the Van Eck sequence # OP VANECK = ( INT n )[]INT: BEGIN [ 1 : IF n < 0 THEN 0 ELSE n FI ]INT result; FOR i TO n DO result[ i ] := 0 OD; [ 0 : UPB result ]INT pos; FOR i FROM 0 TO n DO pos[ i ] := 0 OD; FOR i FROM 2 TO n DO INT j = i - 1; INT prev = result[ j ]; IF pos[ prev ] /= 0 THEN # not a new element # result[ i ] := j - pos[ prev ] FI; pos[ prev ] := j OD; result END # VANECK # ; # construct the first 1000 terms of the sequence # []INT seq = VANECK 1000; # show the first and last 10 elements # FOR i TO 10 DO print( ( " ", whole( seq[ i ], 0 ) ) ) OD; print( ( newline ) ); FOR i FROM UPB seq - 9 TO UPB seq DO print( ( " ", whole( seq[ i ], 0 ) ) ) OD; print( ( newline ) )
END</lang>
- Output:
0 0 1 0 2 0 2 2 1 6 4 7 30 25 67 225 488 0 10 136
ALGOL-M
<lang algol>begin integer array eck[1:1000]; integer i, j;
for i := 1 step 1 until 1000 do
eck[i] := 0;
for i := 1 step 1 until 999 do begin
j := i - 1; while j > 0 and eck[i] <> eck[j] do j := j - 1; if j <> 0 then eck[i+1] := i - j;
end;
for i := 1 step 1 until 10 do
writeon(eck[i]);
write(""); for i := 991 step 1 until 1000 do
writeon(eck[i]);
end</lang>
- Output:
0 0 1 0 2 0 2 2 1 6 4 7 30 25 67 225 488 0 10 136
APL
<lang APL>(10∘↑,[.5]¯10∘↑)(⊢,(⊃∘⌽∊¯1∘↓)∧(1↓⌽)⍳⊃∘⌽)⍣999⊢,0</lang>
- Output:
0 0 1 0 2 0 2 2 1 6 4 7 30 25 67 225 488 0 10 136
AppleScript
Functional
AppleScript is not the tool for the job, but here is a quick assembly from ready-made parts: <lang applescript>use AppleScript version "2.4" use scripting additions
-- vanEck :: Int -> [Int]
on vanEck(n)
-- First n terms of the vanEck sequence. script go on |λ|(xns, i) set {x, ns} to xns set prev to item (1 + x) of ns if 0 ≠ prev then set v to i - prev else set v to 0 end if {{v, insert(ns, x, i)}, v} end |λ| end script {0} & item 2 of mapAccumL(go, ¬ {0, replicate(n, 0)}, enumFromTo(1, n - 1))
end vanEck
TEST ---------------------------
on run
unlines({¬ "First 10 terms:", ¬ showList(vanEck(10)), ¬ "", ¬ "Terms 990 to 1000:", ¬ showList(items -10 thru -1 of vanEck(1000))})
end run
GENERIC --------------------------
-- enumFromTo :: Int -> Int -> [Int] on enumFromTo(m, n)
if m ≤ n then set lst to {} repeat with i from m to n set end of lst to i end repeat lst else {} end if
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
-- insert :: [Int] -> Int -> Int -> [Int]
on insert(xs, i, v)
-- A list updated at position i with value v. set item (1 + i) of xs to v xs
end insert
-- intercalate :: String -> [String] -> String
on intercalate(delim, xs)
set {dlm, my text item delimiters} to ¬ {my text item delimiters, delim} set s to xs as text set my text item delimiters to dlm s
end intercalate
-- map :: (a -> b) -> [a] -> [b]
on map(f, xs)
-- The list obtained by applying f -- to each element of 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
-- mReturn :: First-class m => (a -> b) -> m (a -> b)
on mReturn(f)
-- 2nd class handler function lifted into 1st class script wrapper. if script is class of f then f else script property |λ| : f end script end if
end mReturn
-- '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, i) tell mReturn(f) to set pair to |λ|(item 1 of a, x, i) {item 1 of pair, (item 2 of a) & {item 2 of pair}} end |λ| end script foldl(result, {acc, []}, xs)
end mapAccumL
-- Egyptian multiplication - progressively doubling a list, appending
-- stages of doubling to an accumulator where needed for binary
-- assembly of a target length
-- replicate :: Int -> a -> [a]
on replicate(n, a)
set out to {} if 1 > n then return out set dbl to {a} repeat while (1 < n) if 0 < (n mod 2) then set out to out & dbl set n to (n div 2) set dbl to (dbl & dbl) end repeat return out & dbl
end replicate
-- showList :: [a] -> String
on showList(xs)
"[" & intercalate(", ", map(my str, xs)) & "]"
end showList
-- str :: a -> String
on str(x)
x as string
end str
-- unlines :: [String] -> String
on unlines(xs)
-- A single string formed by the intercalation -- of a list of strings with the newline character. set {dlm, my text item delimiters} to ¬ {my text item delimiters, linefeed} set s to xs as text set my text item delimiters to dlm s
end unlines</lang>
- Output:
First 10 terms: [0, 0, 1, 0, 2, 0, 2, 2, 1, 6] Terms 999 to 1000: [4, 7, 30, 25, 67, 225, 488, 0, 10, 136]
Idiomatic
On the contrary, it's right up AppleScript's street.
<lang applescript>on vanEckSequence(limit)
script o property sequence : {} property lookup : {} end script set term to 0 repeat with i from 1 to (limit - 1) -- 1-based indices. set end of o's sequence to term set t to term + 1 -- 1-based index. repeat (t - (count o's lookup)) times set end of o's lookup to missing value end repeat set previous_i to item t of o's lookup set item t of o's lookup to i if (previous_i is missing value) then set term to 0 else set term to i - previous_i end if end repeat set end of o's sequence to term return o's sequence
end vanEckSequence
-- Task code: tell vanEckSequence(1000) to return {items 1 thru 10, items 991 thru 1000}</lang>
- Output:
<lang applescript>{{0, 0, 1, 0, 2, 0, 2, 2, 1, 6}, {4, 7, 30, 25, 67, 225, 488, 0, 10, 136}}</lang>
ARM Assembly
<lang ARM Assembly> /* ARM assembly Raspberry PI */ /* program vanEckSerie.s */
/* REMARK 1 : this program use routines in a include file see task Include a file language arm assembly for the routine affichageMess conversion10 see at end of this program the instruction include */
/* for constantes see task include a file in arm assembly */ /************************************/ /* Constantes */ /************************************/ .include "../constantes.inc"
.equ MAXI, 1000
/*********************************/ /* Initialized data */ /*********************************/ .data sMessResultElement: .asciz " @ " szCarriageReturn: .asciz "\n"
/*********************************/ /* UnInitialized data */ /*********************************/ .bss sZoneConv: .skip 24 TableVanEck: .skip 4 * MAXI /*********************************/ /* code section */ /*********************************/ .text .global main main: @ entry of program
mov r2,#0 @ begin first element mov r3,#0 @ current counter ldr r4,iAdrTableVanEck @ table address str r2,[r4,r3,lsl #2] @ store first zéro
1: @ begin loop
mov r5,r3 @ init current indice
2:
sub r5,#1 @ decrement cmp r5,#0 @ end table ? movlt r2,#0 @ yes, move zero to next element blt 3f ldr r6,[r4,r5,lsl #2] @ load element cmp r6,r2 @ and compare with the last element bne 2b @ not equal sub r2,r3,r5 @ else compute gap
3:
add r3,r3,#1 @ increment counter str r2,[r4,r3,lsl #2] @ and store new element cmp r3,#MAXI blt 1b mov r2,#0
4: @ loop display ten elements
ldr r0,[r4,r2,lsl #2] ldr r1,iAdrsZoneConv bl conversion10 @ call décimal conversion ldr r0,iAdrsMessResultElement ldr r1,iAdrsZoneConv @ insert conversion in message bl strInsertAtCharInc mov r1,#0 @ final zéro strb r1,[r0,#5] @ bl affichageMess @ display message add r2,#1 @ increment indice cmp r2,#10 @ end ? blt 4b @ no -> loop ldr r0,iAdrszCarriageReturn bl affichageMess mov r2,#MAXI - 10
5: @ loop display ten elements 990-999
ldr r0,[r4,r2,lsl #2] ldr r1,iAdrsZoneConv bl conversion10 @ call décimal conversion ldr r0,iAdrsMessResultElement ldr r1,iAdrsZoneConv @ insert conversion in message bl strInsertAtCharInc mov r1,#0 @ final zéro strb r1,[r0,#5] @ bl affichageMess @ display message add r2,#1 @ increment indice cmp r2,#MAXI @ end ? blt 5b @ no -> loop ldr r0,iAdrszCarriageReturn bl affichageMess
100: @ standard end of the program
mov r0, #0 @ return code mov r7, #EXIT @ request to exit program svc #0 @ perform the system call
iAdrszCarriageReturn: .int szCarriageReturn iAdrsMessResultElement: .int sMessResultElement iAdrsZoneConv: .int sZoneConv iAdrTableVanEck: .int TableVanEck
/***************************************************/ /* ROUTINES INCLUDE */ /***************************************************/ .include "../affichage.inc" </lang>
0 0 1 0 2 0 2 2 1 6 4 7 30 25 67 225 488 0 10 136
AWK
<lang AWK>
- syntax: GAWK -f VAN_ECK_SEQUENCE.AWK
- converted from Go
BEGIN {
limit = 1000 for (i=0; i<limit; i++) { arr[i] = 0 } for (n=0; n<limit-1; n++) { for (m=n-1; m>=0; m--) { if (arr[m] == arr[n]) { arr[n+1] = n - m break } } } printf("terms 1-10:") for (i=0; i<10; i++) { printf(" %d",arr[i]) } printf("\n") printf("terms 991-1000:") for (i=990; i<1000; i++) { printf(" %d",arr[i]) } printf("\n") exit(0)
} </lang>
- Output:
terms 1-10: 0 0 1 0 2 0 2 2 1 6 terms 991-1000: 4 7 30 25 67 225 488 0 10 136
BASIC
<lang basic>10 DEFINT A-Z 20 DIM E(1000) 30 FOR I=0 TO 999 40 FOR J=I-1 TO 0 STEP -1 50 IF E(J)=E(I) THEN E(I+1)=I-J: GOTO 80 60 NEXT J 70 E(I+1)=0 80 NEXT I 90 FOR I=0 TO 9: PRINT E(I);: NEXT 95 PRINT 100 FOR I=990 TO 999: PRINT E(I);: NEXT</lang>
- Output:
0 0 1 0 2 0 2 2 1 6 4 7 30 25 67 225 488 0 10 136
BCPL
<lang bcpl>get "libhdr"
let start() be $( let eck = vec 999
for i = 0 to 999 do eck!i := 0 for i = 0 to 998 do for j = i-1 to 0 by -1 do if eck!i = eck!j then $( eck!(i+1) := i-j break $) for i = 0 to 9 do writed(eck!i, 4) wrch('*N') for i = 990 to 999 do writed(eck!i, 4) wrch('*N')
$)</lang>
- Output:
0 0 1 0 2 0 2 2 1 6 4 7 30 25 67 225 488 0 10 136
C
<lang c>#include <stdlib.h>
- include <stdio.h>
int main(int argc, const char *argv[]) {
const int max = 1000; int *a = malloc(max * sizeof(int)); for (int n = 0; n < max - 1; n ++) { for (int m = n - 1; m >= 0; m --) { if (a[m] == a[n]) { a[n+1] = n - m; break; } } }
printf("The first ten terms of the Van Eck sequence are:\n"); for (int i = 0; i < 10; i ++) printf("%d ", a[i]); printf("\n\nTerms 991 to 1000 of the sequence are:\n"); for (int i = 990; i < 1000; i ++) printf("%d ", a[i]); putchar('\n');
return 0;
}</lang>
- Output:
The first ten terms of the Van Eck sequence are: 0 0 1 0 2 0 2 2 1 6 Terms 991 to 1000 of the sequence are: 4 7 30 25 67 225 488 0 10 136
C#
<lang csharp>using System.Linq; class Program { static void Main() {
int a, b, c, d, e, f, g; int[] h = new int[g = 1000]; for (a = 0, b = 1, c = 2; c < g; a = b, b = c++) for (d = a, e = b - d, f = h[b]; e <= b; e++) if (f == h[d--]) { h[c] = e; break; } void sho(int i) { System.Console.WriteLine(string.Join(" ", h.Skip(i).Take(10))); } sho(0); sho(990); } }</lang>
- Output:
0 0 1 0 2 0 2 2 1 6 4 7 30 25 67 225 488 0 10 136
C++
<lang cpp>#include <iostream>
- include <map>
class van_eck_generator { public:
int next() { int result = last_term; auto iter = last_pos.find(last_term); int next_term = (iter != last_pos.end()) ? index - iter->second : 0; last_pos[last_term] = index; last_term = next_term; ++index; return result; }
private:
int index = 0; int last_term = 0; std::map<int, int> last_pos;
};
int main() {
van_eck_generator gen; int i = 0; std::cout << "First 10 terms of the Van Eck sequence:\n"; for (; i < 10; ++i) std::cout << gen.next() << ' '; for (; i < 990; ++i) gen.next(); std::cout << "\nTerms 991 to 1000 of the sequence:\n"; for (; i < 1000; ++i) std::cout << gen.next() << ' '; std::cout << '\n'; return 0;
}</lang>
- Output:
First 10 terms of the Van Eck sequence: 0 0 1 0 2 0 2 2 1 6 Terms 991 to 1000 of the sequence: 4 7 30 25 67 225 488 0 10 136
Clojure
<lang clojure>(defn van-eck
([] (van-eck 0 0 {})) ([val n seen] (lazy-seq (cons val (let [next (- n (get seen val n))] (van-eck next (inc n) (assoc seen val n)))))))
(println "First 10 terms:" (take 10 (van-eck))) (println "Terms 991 to 1000 terms:" (take 10 (drop 990 (van-eck))))</lang>
- Output:
First 10 terms: (0 0 1 0 2 0 2 2 1 6) Terms 991 to 1000 terms: (4 7 30 25 67 225 488 0 10 136)
Common Lisp
<lang Lisp>
- Tested using CLISP
(defun VanEck (x) (reverse (VanEckh x 0 0 '(0))))
(defun VanEckh (final index curr lst) (if (eq index final) lst (VanEckh final (+ index 1) (howfar curr lst) (cons curr lst))))
(defun howfar (x lst) (howfarh x lst 0))
(defun howfarh (x lst runningtotal) (cond ((null lst) 0) ((eq x (car lst)) (+ runningtotal 1)) (t (howfarh x (cdr lst) (+ runningtotal 1)))))
(format t "The first 10 elements are ~a~%" (VanEck 9)) (format t "The 990-1000th elements are ~a~%" (nthcdr 990 (VanEck 999))) </lang>
- Output:
The first 10 elements are (0 0 1 0 2 0 2 2 1 6) The 990-1000th elements are (4 7 30 25 67 225 488 0 10 136)
<lang Lisp> (defun van-eck-nm-sequence (n m)
(loop with ac repeat m for i = (position (car ac) (cdr ac)) do (push (if i (1+ i) 0) ac) finally (return (nthcdr (1- n) (nreverse ac)))))
(format t "The first 10 elements are: ~{~a ~}~%" (van-eck-nm-sequence 1 10)) (format t "The 991-1000th elements are: ~{~a ~}" (van-eck-nm-sequence 991 1000)) </lang>
- Output:
The first 10 elements are: 0 0 1 0 2 0 2 2 1 6 The 991-1000th elements are: 4 7 30 25 67 225 488 0 10 136
Cowgol
<lang cowgol>include "cowgol.coh";
sub print_list(ptr: [uint16], n: uint8) is
while n > 0 loop print_i16([ptr]); print_char(' '); n := n - 1; ptr := @next ptr; end loop; print_nl();
end sub;
const LIMIT := 1000; var eck: uint16[LIMIT]; MemZero(&eck as [uint8], @bytesof eck); var i: @indexof eck; var j: @indexof eck;
i := 0; while i < LIMIT-1 loop
j := i-1; while j != -1 loop if eck[i] == eck[j] then eck[i+1] := i-j; break; end if; j := j - 1; end loop; i := i + 1;
end loop;
print_list(&eck[0], 10); print_list(&eck[LIMIT-10], 10);</lang>
- Output:
0 0 1 0 2 0 2 2 1 6 4 7 30 25 67 225 488 0 10 136
D
<lang d>import std.stdio;
void vanEck(int firstIndex, int lastIndex) {
int[int] vanEckMap; int last = 0; if (firstIndex == 1) { writefln("VanEck[%d] = %d", 1, 0); } for (int n = 2; n <= lastIndex; n++) { int vanEck = last in vanEckMap ? n - vanEckMap[last] : 0; vanEckMap[last] = n; last = vanEck; if (n >= firstIndex) { writefln("VanEck[%d] = %d", n, vanEck); } }
}
void main() {
writeln("First 10 terms of Van Eck's sequence:"); vanEck(1, 10); writeln; writeln("Terms 991 to 1000 of Van Eck's sequence:"); vanEck(991, 1000);
}</lang>
- Output:
First 10 terms of Van Eck's sequence: VanEck[1] = 0 VanEck[2] = 0 VanEck[3] = 1 VanEck[4] = 0 VanEck[5] = 2 VanEck[6] = 0 VanEck[7] = 2 VanEck[8] = 2 VanEck[9] = 1 VanEck[10] = 6 Terms 991 to 1000 of Van Eck's sequence: VanEck[991] = 4 VanEck[992] = 7 VanEck[993] = 30 VanEck[994] = 25 VanEck[995] = 67 VanEck[996] = 225 VanEck[997] = 488 VanEck[998] = 0 VanEck[999] = 10 VanEck[1000] = 136
Delphi
See Pascal.
Dyalect
<lang dyalect>let max = 1000 var a = Array.empty(max, 0) for n in 0..(max-2) {
var m = n - 1 while m >= 0 { if a[m] == a[n] { a[n+1] = n - m break } m -= 1 }
} print("The first ten terms of the Van Eck sequence are: \(a[0..10])") print("Terms 991 to 1000 of the sequence are: \(a[991..999])")</lang>
- Output:
The first ten terms of the Van Eck sequence are: {0, 0, 1, 0, 2, 0, 2, 2, 1, 6} Terms 991 to 1000 of the sequence are: {7, 30, 25, 67, 225, 488, 0, 10, 136}
F#
The function
<lang fsharp> // Generate Van Eck's Sequence. Nigel Galloway: June 19th., 2019 let ecK()=let n=System.Collections.Generic.Dictionary<int,int>()
Seq.unfold(fun (g,e)->Some(g,((if n.ContainsKey g then let i=n.[g] in n.[g]<-e;e-i else n.[g]<-e;0),e+1)))(0,0)
</lang>
The Task
- First 50
<lang fsharp> ecK() |> Seq.take 50 |> Seq.iter(printf "%d "); printfn "";; </lang>
- Output:
0 0 1 0 2 0 2 2 1 6 0 5 0 2 6 5 4 0 5 3 0 3 2 9 0 4 9 3 6 14 0 6 3 5 15 0 5 3 5 2 17 0 6 11 0 3 8 0 3 3
- 50 from 991
<lang fsharp> ecK() |> Seq.skip 990 |> Seq.take 50|> Seq.iter(printf "%d "); printfn "";; </lang>
- Output:
4 7 30 25 67 225 488 0 10 136 61 0 4 12 72 0 4 4 1 24 41 385 0 7 22 25 22 2 84 68 282 464 0 10 25 9 151 697 0 6 41 20 257 539 0 6 6 1 29 465
- I thought the longest sequence of non zeroes in the first 100 million items might be interesting
It occurs between 32381749 and 32381774:
9 47 47 1 10 33 27 548 548 1 6 33 6 2 154 15657 695734 270964 235721 238076 4896139 655158 7901804 146089 977945 21475977
Alternative recursive procedure
<lang fsharp> open System.Collections.Generic let VanEck() =
let rec _vanEck (num:int) (pos:int) (lastOccurence:Dictionary<int, int>) = match lastOccurence.TryGetValue num with | (true, position) -> set num pos (pos - position) lastOccurence | _ -> set num pos 0 lastOccurence and set num pos next lastOccurenceByNumber = seq { lastOccurenceByNumber.[num] <- pos yield next yield! _vanEck next (pos + 1) lastOccurenceByNumber }
seq { yield 0 yield! _vanEck 0 1 (new Dictionary<int, int>()) }
VanEck() |> Seq.take 10 |> Seq.map (sprintf "%i") |> String.concat " " |> printfn "The first ten terms of the sequence : %s" VanEck() |> Seq.skip 990 |> Seq.take 10 |> Seq.map (sprintf "%i") |> String.concat " " |> printfn "Terms 991 - to - 1000 of the sequence : %s" </lang>
Factor
<lang factor>USING: assocs fry kernel make math namespaces prettyprint sequences ;
- van-eck ( n -- seq )
[ 0 , 1 - H{ } clone '[ building get [ length 1 - ] [ last ] bi _ 3dup 2dup key? [ at - ] [ 3drop 0 ] if , set-at ] times ] { } make ;
1000 van-eck 10 [ head ] [ tail* ] 2bi [ . ] bi@</lang>
- Output:
{ 0 0 1 0 2 0 2 2 1 6 } { 4 7 30 25 67 225 488 0 10 136 }
Fortran
<lang fortran> program VanEck
implicit none integer eck(1000), i, j eck(1) = 0 do 20 i=1, 999 do 10 j=i-1, 1, -1 if (eck(i) .eq. eck(j)) then eck(i+1) = i-j go to 20 end if 10 continue eck(i+1) = 0 20 continue do 30 i=1, 10 30 write (*,'(I4)',advance='no') eck(i) write (*,*) do 40 i=991, 1000 40 write (*,'(I4)',advance='no') eck(i) write (*,*) end program</lang>
- Output:
0 0 1 0 2 0 2 2 1 6 4 7 30 25 67 225 488 0 10 136
FreeBASIC
<lang freebasic> Const limite = 1000
Dim As Integer a(limite), n, m, i
For n = 0 To limite-1
For m = n-1 To 0 Step -1 If a(m) = a(n) Then a(n+1) = n-m: Exit For Next m
Next n
Print "Secuencia de Van Eck:" &Chr(10) Print "Primeros 10 terminos: "; For i = 0 To 9
Print a(i) &" ";
Next i Print Chr(10) & "Terminos 991 al 1000: "; For i = 990 To 999
Print a(i) &" ";
Next i End </lang>
- Output:
Secuencia de Van Eck: Primeros 10 terminos: 0 0 1 0 2 0 2 2 1 6 Terminos 991 al 1000: 4 7 30 25 67 225 488 0 10 136
Fōrmulæ
In this page you can see the solution of this task.
Fōrmulæ programs are not textual, visualization/edition of programs is done showing/manipulating structures but not text (more info). Moreover, there can be multiple visual representations of the same program. Even though it is possible to have textual representation —i.e. XML, JSON— they are intended for transportation effects more than visualization and edition.
The option to show Fōrmulæ programs and their results is showing images. Unfortunately images cannot be uploaded in Rosetta Code.
Go
<lang go>package main
import "fmt"
func main() {
const max = 1000 a := make([]int, max) // all zero by default for n := 0; n < max-1; n++ { for m := n - 1; m >= 0; m-- { if a[m] == a[n] { a[n+1] = n - m break } } } fmt.Println("The first ten terms of the Van Eck sequence are:") fmt.Println(a[:10]) fmt.Println("\nTerms 991 to 1000 of the sequence are:") fmt.Println(a[990:])
}</lang>
- Output:
The first ten terms of the Van Eck sequence are: [0 0 1 0 2 0 2 2 1 6] Terms 991 to 1000 of the sequence are: [4 7 30 25 67 225 488 0 10 136]
Alternatively, using a map to store the latest index of terms previously seen (output as before): <lang go>package main
import "fmt"
func main() {
const max = 1000 a := make([]int, max) // all zero by default seen := make(map[int]int) for n := 0; n < max-1; n++ { if m, ok := seen[a[n]]; ok { a[n+1] = n - m } seen[a[n]] = n } fmt.Println("The first ten terms of the Van Eck sequence are:") fmt.Println(a[:10]) fmt.Println("\nTerms 991 to 1000 of the sequence are:") fmt.Println(a[990:])
}</lang>
Haskell
<lang haskell>import Data.List (elemIndex) import Data.Maybe (maybe)
vanEck :: Int -> [Int] vanEck n = reverse $ iterate go [] !! n
where go [] = [0] go xxs@(x:xs) = maybe 0 succ (elemIndex x xs) : xxs
main :: IO () main = do
print $ vanEck 10 print $ drop 990 (vanEck 1000)</lang>
- Output:
[0,0,1,0,2,0,2,2,1,6] [4,7,30,25,67,225,488,0,10,136]
And if we wanted to look a little further than the 1000th term, we could accumulate a Map of most recently seen positions to improve performance:
<lang haskell>{-# LANGUAGE TupleSections #-}
import qualified Data.Map.Strict as M hiding (drop) import Data.List (mapAccumL) import Data.Maybe (maybe)
vanEck :: [Int] vanEck = 0 : snd (mapAccumL go (0, M.empty) [1 ..])
where go (x, dct) i = ((,) =<< (, M.insert x i dct)) (maybe 0 (i -) (M.lookup x dct))
main :: IO () main =
mapM_ print $ (drop . subtract 10 <*> flip take vanEck) <$> [10, 1000, 10000, 100000, 1000000]</lang>
- Output:
[0,0,1,0,2,0,2,2,1,6] [4,7,30,25,67,225,488,0,10,136] [7,43,190,396,2576,3142,0,7,7,1] [92,893,1125,47187,0,7,34,113,140,2984] [8,86,172,8878,172447,0,6,30,874,34143]
J
The tacit verb (function)
<lang j>VanEck=. (, (<:@:# - }: i: {:))^:(]`0:)</lang>
The output
<lang j> VanEck 9 0 0 1 0 2 0 2 2 1 6
990 }. VanEck 999
4 7 30 25 67 225 488 0 10 136</lang>
A structured derivation of the verb (function)
<lang j> next =. <:@:# - }: i: {: NB. Next term of the sequence VanEck=. (, next)^:(]`0:) f. NB. Appending terms and fixing the verb</lang>
Java
Use map to remember last seen index. Computes each value of the sequence in O(1) time. <lang java> import java.util.HashMap; import java.util.Map;
public class VanEckSequence {
public static void main(String[] args) { System.out.println("First 10 terms of Van Eck's sequence:"); vanEck(1, 10); System.out.println(""); System.out.println("Terms 991 to 1000 of Van Eck's sequence:"); vanEck(991, 1000); } private static void vanEck(int firstIndex, int lastIndex) { Map<Integer,Integer> vanEckMap = new HashMap<>(); int last = 0; if ( firstIndex == 1 ) { System.out.printf("VanEck[%d] = %d%n", 1, 0); } for ( int n = 2 ; n <= lastIndex ; n++ ) { int vanEck = vanEckMap.containsKey(last) ? n - vanEckMap.get(last) : 0; vanEckMap.put(last, n); last = vanEck; if ( n >= firstIndex ) { System.out.printf("VanEck[%d] = %d%n", n, vanEck); } } }
} </lang>
- Output:
First 10 terms of Van Eck's sequence: VanEck[1] = 0 VanEck[2] = 0 VanEck[3] = 1 VanEck[4] = 0 VanEck[5] = 2 VanEck[6] = 0 VanEck[7] = 2 VanEck[8] = 2 VanEck[9] = 1 VanEck[10] = 6 Terms 991 to 1000 of Van Eck's sequence: VanEck[991] = 4 VanEck[992] = 7 VanEck[993] = 30 VanEck[994] = 25 VanEck[995] = 67 VanEck[996] = 225 VanEck[997] = 488 VanEck[998] = 0 VanEck[999] = 10 VanEck[1000] = 136
JavaScript
Either declaratively, without premature optimization:
<lang JavaScript>(() => {
'use strict';
// vanEck :: Int -> [Int] const vanEck = n => reverse( churchNumeral(n)( xs => 0 < xs.length ? cons( maybe( 0, succ, elemIndex(xs[0], xs.slice(1)) ), xs ) : [0] )([]) );
// TEST ----------------------------------------------- const main = () => { console.log('VanEck series:\n') showLog('First 10 terms', vanEck(10)) showLog('Terms 991-1000', vanEck(1000).slice(990)) };
// GENERIC FUNCTIONS ----------------------------------
// Just :: a -> Maybe a const Just = x => ({ type: 'Maybe', Nothing: false, Just: x });
// Nothing :: Maybe a const Nothing = () => ({ type: 'Maybe', Nothing: true, });
// churchNumeral :: Int -> (a -> a) -> a -> a const churchNumeral = n => f => x => Array.from({ length: n }, () => f) .reduce((a, g) => g(a), x)
// cons :: a -> [a] -> [a] const cons = (x, xs) => [x].concat(xs)
// elemIndex :: Eq a => a -> [a] -> Maybe Int const elemIndex = (x, xs) => { const i = xs.indexOf(x); return -1 === i ? ( Nothing() ) : Just(i); };
// maybe :: b -> (a -> b) -> Maybe a -> b const maybe = (v, f, m) => m.Nothing ? v : f(m.Just);
// reverse :: [a] -> [a] const reverse = xs => 'string' !== typeof xs ? ( xs.slice(0).reverse() ) : xs.split().reverse().join();
// showLog :: a -> IO () const showLog = (...args) => console.log( args .map(JSON.stringify) .join(' -> ') );
// succ :: Int -> Int const succ = x => 1 + x;
// MAIN --- return main();
})();</lang>
- Output:
VanEck series: "First 10 terms" -> [0,0,1,0,2,0,2,2,1,6] "Terms 991-1000" -> [4,7,30,25,67,225,488,0,10,136]
or as a map-accumulation, building a look-up table:
<lang javascript>(() => {
'use strict';
// vanEck :: Int -> [Int] const vanEck = n => // First n terms of the vanEck series. [0].concat(mapAccumL( ([x, seen], i) => { const prev = seen[x], v = Boolean(prev) ? ( i - prev ) : 0; return [ [v, (seen[x] = i, seen)], v ]; }, [0, {}],
enumFromTo(1, n - 1) )[1]);
// ----------------------- TEST ------------------------ const main = () => fTable( 'Terms of the VanEck series:\n', n => str(n - 10) + '-' + str(n), xs => JSON.stringify(xs.slice(-10)), vanEck, [10, 1000, 10000] )
// ----------------- GENERIC FUNCTIONS -----------------
// enumFromTo :: Int -> Int -> [Int] const enumFromTo = (m, n) => Array.from({ length: 1 + n - m }, (_, i) => m + i);
// fTable :: String -> (a -> String) -> (b -> String) -> // (a -> b) -> [a] -> String const fTable = (s, xShow, fxShow, f, xs) => { // Heading -> x display function -> // fx display function -> // f -> values -> tabular string const ys = xs.map(xShow), w = Math.max(...ys.map(x => x.length)); return s + '\n' + zipWith( (a, b) => a.padStart(w, ' ') + ' -> ' + b, ys, xs.map(x => fxShow(f(x))) ).join('\n'); };
// Map-accumulation is a combination of map and a catamorphism; // 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.
// 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 [pair[0], a[1].concat(pair[1])]; }, [acc, []]);
// replicate :: Int -> a -> [a] const replicate = (n, x) => Array.from({ length: n }, () => x);
// str :: a -> String const str = x => x.toString();
// zipWith :: (a -> b -> c) -> [a] -> [b] -> [c] const zipWith = (f, xs, ys) => { const lng = Math.min(xs.length, ys.length), as = xs.slice(0, lng), bs = ys.slice(0, lng); return Array.from({ length: lng }, (_, i) => f(as[i], bs[i], i)); };
// MAIN --- return main();
})();</lang>
- Output:
Terms of the VanEck series: 0-10 -> [0,0,1,0,2,0,2,2,1,6] 990-1000 -> [4,7,30,25,67,225,488,0,10,136] 9990-10000 -> [7,43,190,396,2576,3142,0,7,7,1]
jq
<lang># Input: an array
- If the rightmost element, .[-1], does not occur elsewhere, return 0;
- otherwise return the "depth" of its rightmost occurrence in .[0:-2]
def depth:
.[-1] as $x | length as $length | first(range($length-2; -1; -1) as $i | select(.[$i] == $x) | $length - 1 - $i) // 0 ;
- Generate a stream of the first $n van Eck integers:
def vanEck($n):
def v: recurse( if length == $n then empty else . + [depth] end ); [0] | v | .[-1];
- The task:
[vanEck(10)], [vanEck(1000)][990:1001] </lang>
Output
<lang>[0,0,1,0,2,0,2,2,1,6] [4,7,30,25,67,225,488,0,10,136]</lang>
Julia
<lang julia>function vanecksequence(N, startval=0)
ret = zeros(Int, N) ret[1] = startval for i in 1:N-1 lastseen = findlast(x -> x == ret[i], ret[1:i-1]) if lastseen != nothing ret[i + 1] = i - lastseen end end ret
end
println(vanecksequence(10)) println(vanecksequence(1000)[991:1000])
</lang>
- Output:
[0, 0, 1, 0, 2, 0, 2, 2, 1, 6] [4, 7, 30, 25, 67, 225, 488, 0, 10, 136]
Alternate version, with a Dict for memoization (output is the same): <lang julia>function vanecksequence(N, startval=0)
ret = zeros(Int, N) ret[1] = startval lastseen = Dict{Int, Int}() for i in 1:N-1 if haskey(lastseen, ret[i]) ret[i + 1] = i - lastseen[ret[i]] end lastseen[ret[i]] = i end ret
end </lang>
Kotlin
<lang scala>fun main() {
println("First 10 terms of Van Eck's sequence:") vanEck(1, 10) println("") println("Terms 991 to 1000 of Van Eck's sequence:") vanEck(991, 1000)
}
private fun vanEck(firstIndex: Int, lastIndex: Int) {
val vanEckMap = mutableMapOf<Int, Int>() var last = 0 if (firstIndex == 1) { println("VanEck[1] = 0") } for (n in 2..lastIndex) { val vanEck = if (vanEckMap.containsKey(last)) n - vanEckMap[last]!! else 0 vanEckMap[last] = n last = vanEck if (n >= firstIndex) { println("VanEck[$n] = $vanEck") } }
}</lang>
- Output:
First 10 terms of Van Eck's sequence: VanEck[1] = 0 VanEck[2] = 0 VanEck[3] = 1 VanEck[4] = 0 VanEck[5] = 2 VanEck[6] = 0 VanEck[7] = 2 VanEck[8] = 2 VanEck[9] = 1 VanEck[10] = 6 Terms 991 to 1000 of Van Eck's sequence: VanEck[991] = 4 VanEck[992] = 7 VanEck[993] = 30 VanEck[994] = 25 VanEck[995] = 67 VanEck[996] = 225 VanEck[997] = 488 VanEck[998] = 0 VanEck[999] = 10 VanEck[1000] = 136
Lua
<lang Lua>-- Return a table of the first n values of the Van Eck sequence function vanEck (n)
local seq, foundAt = {0} while #seq < n do foundAt = nil for pos = #seq - 1, 1, -1 do if seq[pos] == seq[#seq] then foundAt = pos break end end if foundAt then table.insert(seq, #seq - foundAt) else table.insert(seq, 0) end end return seq
end
-- Show the set of values in table t from key numbers lo to hi function showValues (t, lo, hi)
for i = lo, hi do io.write(t[i] .. " ") end print()
end
-- Main procedure local sequence = vanEck(1000) showValues(sequence, 1, 10) showValues(sequence, 991, 1000)</lang>
- Output:
0 0 1 0 2 0 2 2 1 6 4 7 30 25 67 225 488 0 10 136
MAD
<lang MAD> NORMAL MODE IS INTEGER
DIMENSION E(1000) E(0)=0 THROUGH L1, FOR I=0, 1, I.GE.1000 THROUGH L2, FOR J=I-1, -1, J.L.0 WHENEVER E(J).E.E(I) E(I+1) = I-J TRANSFER TO L1 END OF CONDITIONAL
L2 CONTINUE
E(I+1)=0
L1 CONTINUE
THROUGH S, FOR I=0, 1, I.GE.10
S PRINT FORMAT FMT, I, E(I), I+990, E(I+990)
VECTOR VALUES FMT = $2HE(,I3,2H)=,I3,S5,2HE(,I3,2H)=,I3*$ END OF PROGRAM </lang>
- Output:
E( 0)= 0 E(990)= 4 E( 1)= 0 E(991)= 7 E( 2)= 1 E(992)= 30 E( 3)= 0 E(993)= 25 E( 4)= 2 E(994)= 67 E( 5)= 0 E(995)=225 E( 6)= 2 E(996)=488 E( 7)= 2 E(997)= 0 E( 8)= 1 E(998)= 10 E( 9)= 6 E(999)=136
Mathematica
<lang Mathematica> TakeList[Nest[If[MemberQ[#//Most, #//Last], Join[#, Length[#] - Last@Position[#//Most, #//Last]], Append[#, 0]]&, {0}, 999], {10, -10}] // Column </lang>
- Output:
{0,0,1,0,2,0,2,2,1,6} {4,7,30,25,67,225,488,0,10,136}
Nim
<lang nim>const max = 1000 var a: array[max, int] for n in countup(0, max - 2):
for m in countdown(n - 1, 0): if a[m] == a[n]: a[n + 1] = n - m break
echo "The first ten terms of the Van Eck sequence are:" echo a[..9] echo "\nTerms 991 to 1000 of the sequence are:" echo a[990..^1]</lang>
- Output:
The first ten terms of the Van Eck sequence are: @[0, 0, 1, 0, 2, 0, 2, 2, 1, 6] Terms 991 to 1000 of the sequence are: @[4, 7, 30, 25, 67, 225, 488, 0, 10, 136]
Pascal
I memorize the last position of each number that occured and use a circular buffer to remember last values. Running once through the list of last positions maybe faster Try it online! takes only 1.4 s for 32,381,775 <lang pascal>program VanEck; {
- A: The first term is zero.
Repeatedly apply: If the last term is *new* to the sequence so far then:
B: The next term is zero.
Otherwise:
C: The next term is how far back this last term occured previousely.} uses
sysutils;
const
MAXNUM = 32381775;//1000*1000*1000; MAXSEENIDX = (1 shl 7)-1;
var
PosBefore : array of UInt32; LastSeen : array[0..MAXSEENIDX]of UInt32;// circular buffer SeenIdx,HaveSeen : Uint32;
procedure OutSeen(Cnt:NativeInt); var
I,S_Idx : NativeInt;
Begin
IF Cnt > MAXSEENIDX then Cnt := MAXSEENIDX; If Cnt > HaveSeen then Cnt := HaveSeen; S_Idx := SeenIdx; S_Idx := (S_Idx-Cnt); IF S_Idx < 0 then inc(S_Idx,MAXSEENIDX); For i := 1 to Cnt do Begin write(' ',LastSeen[S_Idx]); S_Idx:= (S_Idx+1) AND MAXSEENIDX; end; writeln;
end;
procedure Test(MaxTestCnt: Uint32); var
i, actnum, Posi, S_Idx: Uint32; {$IFDEF FPC} pPosBef, pSeen: pUint32; {$ELSE} pPosBef, pSeen: array of UInt32; {$ENDIF}
begin
HaveSeen := 0; if MaxTestCnt > MAXNUM then EXIT;
Fillchar(LastSeen, SizeOf(LastSeen), #0); //setlength and clear setlength(PosBefore, 0); setlength(PosBefore, MaxTestCnt);
{$IFDEF FPC} pPosBef := @PosBefore[0]; pSeen := @LastSeen[0]; {$ELSE} SetLength(pSeen, SizeOf(LastSeen)); setlength(pPosBef, MaxTestCnt); move(PosBefore[0], pPosBef[0], length(pPosBef)); move(LastSeen[0], pSeen[0], length(pSeen)); {$ENDIF}
S_Idx := 0; i := 1; actnum := 0; repeat // save value pSeen[S_Idx] := actnum; S_Idx := (S_Idx + 1) and MAXSEENIDX; //examine new value often out of cache Posi := pPosBef[actnum]; pPosBef[actnum] := i;
// if Posi=0 ? actnum = 0:actnum = i-Posi
if Posi = 0 then actnum := 0 else actnum := i - Posi; inc(i); until i > MaxTestCnt; HaveSeen := i - 1; SeenIdx := S_Idx;
{$IFNDEF FPC}
move(pPosBef[0], PosBefore[0], length(pPosBef)); move(pSeen[0], LastSeen[0], length(pSeen)); {$ENDIF}
end;
Begin
Test(10) ; OutSeen(10000); Test(1000); OutSeen(10); Test(MAXNUM); OutSeen(28); setlength(PosBefore,0);
end.</lang>
- Output:
0 0 1 0 2 0 2 2 1 6 4 7 30 25 67 225 488 0 10 136 0 9 47 47 1 10 33 27 548 548 1 6 33 6 2 154 15657 695734 270964 235721 238076 4896139 655158 7901804 146089 977945 21475977 0
Perl
<lang perl>use strict; use warnings; use feature 'say';
sub van_eck {
my($init,$max) = @_; my(%v,$k); my @V = my $i = $init; for (1..$max) { $k++; my $t = $v{$i} ? $k - $v{$i} : 0; $v{$i} = $k; push @V, $i = $t; } @V;
}
for (
['A181391', 0], ['A171911', 1], ['A171912', 2], ['A171913', 3], ['A171914', 4], ['A171915', 5], ['A171916', 6], ['A171917', 7], ['A171918', 8],
) {
my($seq, $start) = @$_; my @seq = van_eck($start,1000); say <<~"END"; Van Eck sequence OEIS:$seq; with the first term: $start First 10 terms: @{[@seq[0 .. 9]]} Terms 991 through 1000: @{[@seq[990..999]]} END
}</lang>
- Output:
Van Eck sequence OEIS:A181391; with the first term: 0 First 10 terms: 0 0 1 0 2 0 2 2 1 6 Terms 991 through 1000: 4 7 30 25 67 225 488 0 10 136 Van Eck sequence OEIS:A171911; with the first term: 1 First 10 terms: 1 0 0 1 3 0 3 2 0 3 Terms 991 through 1000: 0 6 53 114 302 0 5 9 22 71 Van Eck sequence OEIS:A171912; with the first term: 2 First 10 terms: 2 0 0 1 0 2 5 0 3 0 Terms 991 through 1000: 8 92 186 0 5 19 41 413 0 5 Van Eck sequence OEIS:A171913; with the first term: 3 First 10 terms: 3 0 0 1 0 2 0 2 2 1 Terms 991 through 1000: 5 5 1 17 192 0 6 34 38 179 Van Eck sequence OEIS:A171914; with the first term: 4 First 10 terms: 4 0 0 1 0 2 0 2 2 1 Terms 991 through 1000: 33 410 0 6 149 0 3 267 0 3 Van Eck sequence OEIS:A171915; with the first term: 5 First 10 terms: 5 0 0 1 0 2 0 2 2 1 Terms 991 through 1000: 60 459 0 7 13 243 0 4 10 211 Van Eck sequence OEIS:A171916; with the first term: 6 First 10 terms: 6 0 0 1 0 2 0 2 2 1 Terms 991 through 1000: 6 19 11 59 292 0 6 6 1 12 Van Eck sequence OEIS:A171917; with the first term: 7 First 10 terms: 7 0 0 1 0 2 0 2 2 1 Terms 991 through 1000: 11 7 2 7 2 2 1 34 24 238 Van Eck sequence OEIS:A171918; with the first term: 8 First 10 terms: 8 0 0 1 0 2 0 2 2 1 Terms 991 through 1000: 16 183 0 6 21 10 249 0 5 48
Phix
Just like the pascal entry, instead of searching/dictionaries use a fast direct/parallel lookup table,
and likewise this can easily create a 32-million-long table in under 2s.
While dictionaries are pretty fast, there is a huge overhead adding/updating millions of entries compared to a flat list of int.
constant lim = 1000 sequence van_eck = repeat(0,lim), pos_before = repeat(0,lim) for n=1 to lim-1 do integer vn = van_eck[n]+1, prev = pos_before[vn] if prev!=0 then van_eck[n+1] = n - prev end if pos_before[vn] = n end for printf(1,"The first ten terms of the Van Eck sequence are:%v\n",{van_eck[1..10]}) printf(1,"Terms 991 to 1000 of the sequence are:%V\n",{van_eck[991..1000]})
- Output:
The first ten terms of the Van Eck sequence are:{0,0,1,0,2,0,2,2,1,6} Terms 991 to 1000 of the sequence are:{4,7,30,25,67,225,488,0,10,136}
PL/M
<lang plm>100H: BDOS: PROCEDURE (FN, ARG); DECLARE FN BYTE, ARG ADDRESS; GO TO 5; END BDOS; EXIT: PROCEDURE; CALL BDOS(0,0); END EXIT; PRINT: PROCEDURE (S); DECLARE S ADDRESS; CALL BDOS(9,S); END PRINT;
PRINT$NUMBER: PROCEDURE (N);
DECLARE S (7) BYTE INITIAL ('..... $'); DECLARE (N, P) ADDRESS, C BASED P BYTE; P = .S(5);
DIGIT:
P = P - 1; C = N MOD 10 + '0'; N = N / 10; IF N > 0 THEN GO TO DIGIT; CALL PRINT(P);
END PRINT$NUMBER;
PRINT$SLICE: PROCEDURE (LIST, N);
DECLARE (I, N, LIST, L BASED LIST) ADDRESS; DO I=0 TO N-1; CALL PRINT$NUMBER(L(I)); END; CALL PRINT(.(13,10,'$'));
END PRINT$SLICE;
DECLARE ECK (1000) ADDRESS; DECLARE (I, J) ADDRESS;
ECK(0) = 0; DO I=0 TO LAST(ECK)-1;
J = I - 1; DO WHILE J <> 0FFFFH; /* WHAT IS SIGNED MATH */ IF ECK(I) = ECK(J) THEN DO; ECK(I+1) = I-J; GO TO NEXT; END; J = J - 1; END; ECK(I+1) = 0;
NEXT: END;
CALL PRINT$SLICE(.ECK(0), 10); CALL PRINT$SLICE(.ECK(990), 10); CALL EXIT; EOF</lang>
- Output:
0 0 1 0 2 0 2 2 1 6 4 7 30 25 67 225 488 0 10 136
Prolog
<lang prolog>van_eck_init(v(0, 0, _assoc)):-
empty_assoc(_assoc).
van_eck_next(v(Index, Last_term, Last_pos), v(Index1, Next_term, Last_pos1)):-
(get_assoc(Last_term, Last_pos, V) -> Next_term is Index - V ; Next_term = 0 ), Index1 is Index + 1, put_assoc(Last_term, Last_pos, Index, Last_pos1).
van_eck_sequence(N, Seq):-
van_eck_init(V), van_eck_sequence(N, V, Seq).
van_eck_sequence(0, _, []):-!. van_eck_sequence(N, V, [Term|Rest]):-
V = v(_, Term, _), van_eck_next(V, V1), N1 is N - 1, van_eck_sequence(N1, V1, Rest).
write_list(From, To, _, _):-
To < From, !.
write_list(_, _, _, []):-!. write_list(From, To, N, [_|Rest]):-
From > N, !, N1 is N + 1, write_list(From, To, N1, Rest).
write_list(From, To, N, [E|Rest]):-
writef('%t ', [E]), F1 is From + 1, N1 is N + 1, write_list(F1, To, N1, Rest).
write_list(From, To, List):-
write_list(From, To, 1, List), nl.
main:-
van_eck_sequence(1000, Seq), writeln('First 10 terms of the Van Eck sequence:'), write_list(1, 10, Seq), writeln('Terms 991 to 1000 of the Van Eck sequence:'), write_list(991, 1000, Seq).</lang>
- Output:
First 10 terms of the Van Eck sequence: 0 0 1 0 2 0 2 2 1 6 Terms 991 to 1000 of the Van Eck sequence: 4 7 30 25 67 225 488 0 10 136
Python
Python: Using a dict
<lang python>def van_eck():
n, seen, val = 0, {}, 0 while True: yield val last = {val: n} val = n - seen.get(val, n) seen.update(last) n += 1
- %%
if __name__ == '__main__':
print("Van Eck: first 10 terms: ", list(islice(van_eck(), 10))) print("Van Eck: terms 991 - 1000:", list(islice(van_eck(), 1000))[-10:])</lang>
- Output:
Van Eck: first 10 terms: [0, 0, 1, 0, 2, 0, 2, 2, 1, 6] Van Eck: terms 991 - 1000: [4, 7, 30, 25, 67, 225, 488, 0, 10, 136]
Python: List based
The following alternative stores the sequence so far in a list seen
rather than the first example that just stores last occurrences in a dict.
<lang python>def van_eck():
n = 0 seen = [0] val = 0 while True: yield val if val in seen[1:]: val = seen.index(val, 1) else: val = 0 seen.insert(0, val) n += 1</lang>
- Output:
As before.
Python: Composition of pure functions
As an alternative to the use of generators, a declarative definition in terms of a Church numeral function:
<lang python>Van Eck sequence
from functools import reduce from itertools import repeat
- vanEck :: Int -> [Int]
def vanEck(n):
First n terms of the van Eck sequence.
return churchNumeral(n)( lambda xs: cons( maybe(0)(succ)( elemIndex(xs[0])(xs[1:]) ) )(xs) if xs else [0] )([])[::-1]
- TEST ----------------------------------------------------
def main():
Terms of the Van Eck sequence print( main.__doc__ + ':\n\n' + 'First 10: '.rjust(18, ' ') + repr(vanEck(10)) + '\n' + '991 - 1000: '.rjust(18, ' ') + repr(vanEck(1000)[990:]) )
- GENERIC -------------------------------------------------
- Just :: a -> Maybe a
def Just(x):
Constructor for an inhabited Maybe (option type) value. Wrapper containing the result of a computation. return {'type': 'Maybe', 'Nothing': False, 'Just': x}
- Nothing :: Maybe a
def Nothing():
Constructor for an empty Maybe (option type) value. Empty wrapper returned where a computation is not possible. return {'type': 'Maybe', 'Nothing': True}
- churchNumeral :: Int -> (a -> a) -> a -> a
def churchNumeral(n):
n applications of a function return lambda f: lambda x: reduce( lambda a, g: g(a), repeat(f, n), x )
- cons :: a -> [a] -> [a]
def cons(x):
Construction of a list from a head and a tail. return lambda xs: [x] + xs
- elemIndex :: Eq a => a -> [a] -> Maybe Int
def elemIndex(x):
Just the index of the first element in xs which is equal to x, or Nothing if there is no such element. def go(xs): try: return Just(xs.index(x)) except ValueError: return Nothing() return go
- maybe :: b -> (a -> b) -> Maybe a -> b
def maybe(v):
Either the default value v, if m is Nothing, or the application of f to x, where m is Just(x). return lambda f: lambda m: v if None is m or m.get('Nothing') else ( f(m.get('Just')) )
- succ :: Enum a => a -> a
def succ(x):
The successor of a value. For numeric types, (1 +). return 1 + x
- MAIN ---
if __name__ == '__main__':
main()</lang>
- Output:
Terms of the Van Eck sequence: First 10: [0, 0, 1, 0, 2, 0, 2, 2, 1, 6] 991 - 1000: [4, 7, 30, 25, 67, 225, 488, 0, 10, 136]
Or if we lose sight, for a moment, of the good advice of Donald Knuth, and fall into optimising more than is needed for the first 1000 terms, then we can define the vanEck series as a map accumulation over a range, with an array of positions as the accumulator.
<lang python>Van Eck series by map-accumulation
from functools import reduce from itertools import repeat
- vanEck :: Int -> [Int]
def vanEck(n):
First n terms of the vanEck sequence. def go(xns, i): x, ns = xns
prev = ns[x] v = i - prev if 0 is not prev else 0 return ( (v, insert(ns, x, i)), v )
return [0] + mapAccumL(go)((0, list(repeat(0, n))))( range(1, n) )[1]
- -------------------------- TEST --------------------------
- main :: IO ()
def main():
The last 10 of the first N vanEck terms print( fTable(main.__doc__ + ':\n')( lambda m: 'N=' + str(m), repr, lambda n: vanEck(n)[-10:], [10, 1000, 10000] ) )
- ----------------------- FORMATTING -----------------------
- fTable :: String -> (a -> String) ->
- (b -> String) -> (a -> b) -> [a] -> String
def fTable(s):
Heading -> x display function -> fx display function -> f -> xs -> tabular string. def go(xShow, fxShow, f, xs): ys = [xShow(x) for x in xs] w = max(map(len, ys)) return s + '\n' + '\n'.join(map( lambda x, y: y.rjust(w, ' ') + ' -> ' + fxShow(f(x)), xs, ys )) return go
- ------------------------ GENERIC -------------------------
- insert :: Array Int -> Int -> Int -> Array Int
def insert(xs, i, v):
An array updated at position i with value v. xs[i] = v return xs
- mapAccumL :: (acc -> x -> (acc, y)) -> acc -> [x] -> (acc, [y])
def mapAccumL(f):
A tuple of an accumulation and a list derived by a combined map and fold, with accumulation from left to right. def go(a, x): tpl = f(a[0], x) return (tpl[0], a[1] + [tpl[1]]) return lambda acc: lambda xs: ( reduce(go, xs, (acc, [])) )
- MAIN ---
if __name__ == '__main__':
main()</lang>
- Output:
The last 10 of the first N vanEck terms: N=10 -> [0, 0, 1, 0, 2, 0, 2, 2, 1, 6] N=1000 -> [4, 7, 30, 25, 67, 225, 488, 0, 10, 136] N=10000 -> [7, 43, 190, 396, 2576, 3142, 0, 7, 7, 1]
Racket
<lang racket>#lang racket (require racket/stream)
(define (van-eck)
(define (next val n seen) (define val1 (- n (hash-ref seen val n))) (stream-cons val (next val1 (+ n 1) (hash-set seen val n)))) (next 0 0 (hash)))
(define (get m n s)
(stream->list (stream-take (stream-tail s m) (- n m))))
"First 10 terms:" (get 0 10 (van-eck)) "Terms 991 to 1000 terms:" (get 990 1000 (van-eck)) ; counting from 0</lang>
- Output:
"First 10 terms:" (0 0 1 0 2 0 2 2 1 6) "Terms 991 to 1000 terms:" (4 7 30 25 67 225 488 0 10 136)
Raku
(formerly Perl 6) There is not a Van Eck sequence, rather a series of related sequences that differ in their starting value. This task is nominally for the sequence starting with the value 0. This Raku implementation will handle any integer starting value.
Specifically handles:
- OEIS:A181391 - Van Eck sequence starting with 0
- OEIS:A171911 - Van Eck sequence starting with 1
- OEIS:A171912 - Van Eck sequence starting with 2
- OEIS:A171913 - Van Eck sequence starting with 3
- OEIS:A171914 - Van Eck sequence starting with 4
- OEIS:A171915 - Van Eck sequence starting with 5
- OEIS:A171916 - Van Eck sequence starting with 6
- OEIS:A171917 - Van Eck sequence starting with 7
- OEIS:A171918 - Van Eck sequence starting with 8
among others.
Implemented as lazy, extendable lists.
<lang perl6>sub n-van-ecks ($init) {
$init, -> $i, { state %v; state $k; $k++; my $t = %v{$i}.defined ?? $k - %v{$i} !! 0; %v{$i} = $k; $t } ... *
}
for <
A181391 0 A171911 1 A171912 2 A171913 3 A171914 4 A171915 5 A171916 6 A171917 7 A171918 8
> -> $seq, $start {
my @seq = n-van-ecks($start);
# The task put qq:to/END/
Van Eck sequence OEIS:$seq; with the first term: $start First 10 terms: {@seq[^10]} Terms 991 through 1000: {@seq[990..999]} END
}</lang>
- Output:
Van Eck sequence OEIS:A181391; with the first term: 0 First 10 terms: 0 0 1 0 2 0 2 2 1 6 Terms 991 through 1000: 4 7 30 25 67 225 488 0 10 136 Van Eck sequence OEIS:A171911; with the first term: 1 First 10 terms: 1 0 0 1 3 0 3 2 0 3 Terms 991 through 1000: 0 6 53 114 302 0 5 9 22 71 Van Eck sequence OEIS:A171912; with the first term: 2 First 10 terms: 2 0 0 1 0 2 5 0 3 0 Terms 991 through 1000: 8 92 186 0 5 19 41 413 0 5 Van Eck sequence OEIS:A171913; with the first term: 3 First 10 terms: 3 0 0 1 0 2 0 2 2 1 Terms 991 through 1000: 5 5 1 17 192 0 6 34 38 179 Van Eck sequence OEIS:A171914; with the first term: 4 First 10 terms: 4 0 0 1 0 2 0 2 2 1 Terms 991 through 1000: 33 410 0 6 149 0 3 267 0 3 Van Eck sequence OEIS:A171915; with the first term: 5 First 10 terms: 5 0 0 1 0 2 0 2 2 1 Terms 991 through 1000: 60 459 0 7 13 243 0 4 10 211 Van Eck sequence OEIS:A171916; with the first term: 6 First 10 terms: 6 0 0 1 0 2 0 2 2 1 Terms 991 through 1000: 6 19 11 59 292 0 6 6 1 12 Van Eck sequence OEIS:A171917; with the first term: 7 First 10 terms: 7 0 0 1 0 2 0 2 2 1 Terms 991 through 1000: 11 7 2 7 2 2 1 34 24 238 Van Eck sequence OEIS:A171918; with the first term: 8 First 10 terms: 8 0 0 1 0 2 0 2 2 1 Terms 991 through 1000: 16 183 0 6 21 10 249 0 5 48
REXX
using a list
This REXX version allows the specification of the start and end of the Van Eck sequence (to be displayed) as
well as the initial starting element (the default is zero).
<lang rexx>/*REXX pgm generates/displays the 'start ──► end' elements of the Van Eck sequence.*/
parse arg LO HI $ . /*obtain optional arguments from the CL*/
if LO== | LO=="," then LO= 1 /*Not specified? Then use the default.*/
if HI== | HI=="," then HI= 10 /* " " " " " " */
if $== | $=="," then $= 0 /* " " " " " " */
$$=; z= $ /*$$: old seq: $: initial value of seq*/
do HI-1; z= wordpos( reverse(z), reverse($$) ); $$= $; $= $ z end /*HI-1*/ /*REVERSE allows backwards search in $.*/ /*stick a fork in it, we're all done. */
say 'terms ' LO " through " HI ' of the Van Eck sequence are: ' subword($,LO,HI-LO+1)</lang>
- output when using the default inputs:
terms 1 through 10 of the Van Eck sequence are: 0 0 1 0 2 0 2 2 1 6
- output when using the inputs of: 991 1000
terms 991 through 1000 of the Van Eck sequence are: 4 7 30 25 67 225 488 0 10 136
- output when using the inputs of: 1 20 6
terms 1 through 20 of the Van Eck sequence are: 6 0 0 1 0 2 0 2 2 1 6 10 0 6 3 0 3 2 9 0
using a dictionary
This REXX version (which uses a dictionary) is about 20,000 times faster (when using larger numbers) than
using a list (in finding the previous location of an "old" number (term).
<lang rexx>/*REXX pgm generates/displays the 'start ──► end' elements of the Van Eck sequence.*/
parse arg LO HI sta . /*obtain optional arguments from the CL*/
if LO== | LO=="," then LO= 1 /*Not specified? Then use the default.*/
if HI== | HI=="," then HI= 10 /* " " " " " " */
if sta== | sta=="," then sta= 0 /* " " " " " " */
$.0= sta; x= sta; @.=. /*$.: the Van Eck sequence as a list.*/
do #=1 for HI-1 /*X: is the last term being examined. */ if @.x==. then do; @.x= #; $.#= 0; x= 0; end /*new term.*/ else do; z= # - @.x; $.#= z; @.x= #; x= z; end /*old term.*/ end /*#*/ /*Z: the new term being added to list.*/ LOw= LO - 1; out= $.LOw /*initialize the output value. */ do j=LO to HI-1; out= out $.j /*build a list for the output display. */ end /*j*/ /*stick a fork in it, we're all done. */
say 'terms ' LO " through " HI ' of the Van Eck sequence are: ' out</lang>
- output is identical to the 1st REXX version.
Ruby
Ruby: Using an Array
<lang ruby>van_eck = Enumerator.new do |y|
ar = [0] loop do y << (term = ar.last) # yield ar << (ar.count(term)==1 ? 0 : ar.size - 1 - ar[0..-2].rindex(term)) end
end
ve = van_eck.take(1000) p ve.first(10), ve.last(10) </lang>
- Output:
[0, 0, 1, 0, 2, 0, 2, 2, 1, 6] [4, 7, 30, 25, 67, 225, 488, 0, 10, 136]
Ruby: Using a Hash
<lang ruby>class VenEch
include Enumerable def initialize() @i = 0; @val = 0; @seen = {}; end
def add_num num @val = @i - @seen.fetch(num, @i) @seen[num] = @i @i += 1 end
def each(&block) loop { block.call(@val); add_num @val } end
end
ve = VenEch.new.take(1000) p ve.first(10), ve.last(10) </lang>
- Output:
[0, 0, 1, 0, 2, 0, 2, 2, 1, 6] [4, 7, 30, 25, 67, 225, 488, 0, 10, 136]
Rust
<lang rust>fn van_eck_sequence() -> impl std::iter::Iterator<Item = i32> {
let mut index = 0; let mut last_term = 0; let mut last_pos = std::collections::HashMap::new(); std::iter::from_fn(move || { let result = last_term; let mut next_term = 0; if let Some(v) = last_pos.get_mut(&last_term) { next_term = index - *v; *v = index; } else { last_pos.insert(last_term, index); } last_term = next_term; index += 1; Some(result) })
}
fn main() {
let mut v = van_eck_sequence().take(1000); println!("First 10 terms of the Van Eck sequence:"); for n in v.by_ref().take(10) { print!("{} ", n); } println!("\nTerms 991 to 1000 of the Van Eck sequence:"); for n in v.skip(980) { print!("{} ", n); } println!();
}</lang>
- Output:
First 10 terms of the Van Eck sequence: 0 0 1 0 2 0 2 2 1 6 Terms 991 to 1000 of the Van Eck sequence: 4 7 30 25 67 225 488 0 10 136
Scala
<lang scala> object VanEck extends App {
def vanEck(n: Int): List[Int] = {
def vanEck(values: List[Int]): List[Int] = if (values.size < n) vanEck(math.max(0, values.indexOf(values.head, 1)) :: values) else values
vanEck(List(0)).reverse }
val vanEck1000 = vanEck(1000) println(s"The first 10 terms are ${vanEck1000.take(10)}.") println(s"Terms 991 to 1000 are ${vanEck1000.drop(990)}.")
} </lang>
- Output:
The first 10 terms are List(0, 0, 1, 0, 2, 0, 2, 2, 1, 6). Terms 991 to 1000 are List(4, 7, 30, 25, 67, 225, 488, 0, 10, 136).
SNOBOL4
<lang snobol4> define('eck(n)i,j') :(eck_end) eck eck = array(n,0)
i = 0
eouter i = lt(i,n - 1) i + 1 :f(return)
j = i
einner j = gt(j,0) j - 1 :f(eouter)
eck = eq(eck,eck<j>) i - j :s(eouter)f(einner)
eck_end
define('list(arr,start,stop)') :(list_end)
list list = list arr<start> ' '
start = lt(start,stop) start + 1 :s(list)f(return)
list_end
ecks = eck(1000) output = list(ecks, 1, 10) output = list(ecks, 991, 1000)
end</lang>
- Output:
0 0 1 0 2 0 2 2 1 6 4 7 30 25 67 225 488 0 10 136
Sidef
<lang ruby>func van_eck(n) {
var seen = Hash() var seq = [0] var prev = seq[-1]
for k in (1 ..^ n) { seq << (seen.has(prev) ? (k - seen{prev}) : 0) seen{prev} = k prev = seq[-1] }
seq
}
say van_eck(10) say van_eck(1000).slice(991-1, 1000-1)</lang>
- Output:
[0, 0, 1, 0, 2, 0, 2, 2, 1, 6] [4, 7, 30, 25, 67, 225, 488, 0, 10, 136]
Swift
<lang swift>struct VanEckSequence: Sequence, IteratorProtocol {
private var index = 0 private var lastTerm = 0 private var lastPos = Dictionary<Int, Int>() mutating func next() -> Int? { let result = lastTerm var nextTerm = 0 if let v = lastPos[lastTerm] { nextTerm = index - v } lastPos[lastTerm] = index lastTerm = nextTerm index += 1 return result }
}
let seq = VanEckSequence().prefix(1000)
print("First 10 terms of the Van Eck sequence:") for n in seq.prefix(10) {
print(n, terminator: " ")
} print("\nTerms 991 to 1000 of the Van Eck sequence:") for n in seq.dropFirst(990) {
print(n, terminator: " ")
} print()</lang>
- Output:
First 10 terms of the Van Eck sequence: 0 0 1 0 2 0 2 2 1 6 Terms 991 to 1000 of the Van Eck sequence: 4 7 30 25 67 225 488 0 10 136
Tcl
<lang Tcl>## Mathematically, the first term has index "0", not "1". We do that, also.
set ::vE 0
proc vanEck {n} {
global vE vEocc while {$n >= [set k [expr {[llength $vE] - 1}]]} { set kv [lindex $vE $k] ## value $kv @ $k is not yet stuffed into vEocc() lappend vE [expr {[info exists vEocc($kv)] ? $k - $vEocc($kv) : 0}] set vEocc($kv) $k } return [lindex $vE $n]
}
proc show {func from to} {
for {set n $from} {$n <= $to} {incr n} { append r " " [$func $n] } puts "${func}($from..$to) =$r"
}
show vanEck 0 9 show vanEck 990 999</lang>
- Output:
vanEck(0..9) = 0 0 1 0 2 0 2 2 1 6 vanEck(990..999) = 4 7 30 25 67 225 488 0 10 136
Visual Basic .NET
<lang vbnet>Imports System.Linq Module Module1
Dim h() As Integer Sub sho(i As Integer) Console.WriteLine(String.Join(" ", h.Skip(i).Take(10))) End Sub Sub Main() Dim a, b, c, d, f, g As Integer : g = 1000 h = new Integer(g){} : a = 0 : b = 1 : For c = 2 To g f = h(b) : For d = a To 0 Step -1 If f = h(d) Then h(c) = b - d: Exit For Next : a = b : b = c : Next : sho(0) : sho(990) End Sub
End Module </lang>
- Output:
Same as C#.
Wren
<lang javascript>var max = 1000 var a = List.filled(max, 0) var seen = {} for (n in 0...max-1) {
var m = seen[a[n]] if (m != null) a[n+1] = n - m seen[a[n]] = n
} System.print("The first ten terms of the Van Eck sequence are:") System.print(a[0...10]) System.print("\nTerms 991 to 1000 of the sequence are:") System.print(a[990..-1])</lang>
- Output:
The first ten terms of the Van Eck sequence are: [0, 0, 1, 0, 2, 0, 2, 2, 1, 6] Terms 991 to 1000 of the sequence are: [4, 7, 30, 25, 67, 225, 488, 0, 10, 136]
zkl
<lang zkl>fcn vanEck(startAt=0){ // --> iterator
(startAt).walker(*).tweak(fcn(n,seen,rprev){ prev,t := rprev.value, n - seen.find(prev,n); seen[prev] = n; rprev.set(t); t }.fp1(Dictionary(),Ref(startAt))).push(startAt)
}</lang> <lang zkl>foreach n in (9){
ve:=vanEck(n); println("The first ten terms of the Van Eck (%d) sequence are:".fmt(n)); println("\t",ve.walk(10).concat(",")); println(" Terms 991 to 1000 of the sequence are:"); println("\t",ve.drop(990-10).walk(10).concat(","));
}</lang>
- Output:
The first ten terms of the Van Eck (0) sequence are: 0,0,1,0,2,0,2,2,1,6 Terms 991 to 1000 of the sequence are: 4,7,30,25,67,225,488,0,10,136 The first ten terms of the Van Eck (1) sequence are: 1,0,0,1,3,0,3,2,0,3 Terms 991 to 1000 of the sequence are: 0,6,53,114,302,0,5,9,22,71 The first ten terms of the Van Eck (2) sequence are: 2,0,0,1,0,2,5,0,3,0 Terms 991 to 1000 of the sequence are: 8,92,186,0,5,19,41,413,0,5 The first ten terms of the Van Eck (3) sequence are: 3,0,0,1,0,2,0,2,2,1 Terms 991 to 1000 of the sequence are: 5,5,1,17,192,0,6,34,38,179 The first ten terms of the Van Eck (4) sequence are: 4,0,0,1,0,2,0,2,2,1 Terms 991 to 1000 of the sequence are: 33,410,0,6,149,0,3,267,0,3 The first ten terms of the Van Eck (5) sequence are: 5,0,0,1,0,2,0,2,2,1 Terms 991 to 1000 of the sequence are: 60,459,0,7,13,243,0,4,10,211 The first ten terms of the Van Eck (6) sequence are: 6,0,0,1,0,2,0,2,2,1 Terms 991 to 1000 of the sequence are: 6,19,11,59,292,0,6,6,1,12 The first ten terms of the Van Eck (7) sequence are: 7,0,0,1,0,2,0,2,2,1 Terms 991 to 1000 of the sequence are: 11,7,2,7,2,2,1,34,24,238 The first ten terms of the Van Eck (8) sequence are: 8,0,0,1,0,2,0,2,2,1 Terms 991 to 1000 of the sequence are: 16,183,0,6,21,10,249,0,5,48
- Programming Tasks
- Solutions by Programming Task
- 11l
- 8080 Assembly
- 8086 Assembly
- AArch64 Assembly
- Ada
- ALGOL 68
- ALGOL-M
- APL
- AppleScript
- ARM Assembly
- AWK
- BASIC
- BCPL
- C
- C sharp
- C++
- Clojure
- Common Lisp
- Cowgol
- D
- Delphi
- Dyalect
- F Sharp
- Factor
- Fortran
- FreeBASIC
- Fōrmulæ
- Go
- Haskell
- J
- Java
- JavaScript
- Jq
- Julia
- Kotlin
- Lua
- MAD
- Mathematica
- Nim
- Pascal
- Perl
- Phix
- PL/M
- Prolog
- Python
- Racket
- Raku
- REXX
- Ruby
- Rust
- Scala
- SNOBOL4
- Sidef
- Swift
- Tcl
- Visual Basic .NET
- Wren
- Zkl