Data update

This commit is contained in:
Ingy döt Net 2026-04-30 12:34:36 -04:00
parent 4bb20c9b71
commit cbaf4c4b64
12390 changed files with 318560 additions and 27248 deletions

View file

@ -0,0 +1,56 @@
with Ada.Text_IO, Ada.Integer_Text_IO, Ada.Strings.Unbounded,
Ada.Strings.Unbounded.Text_IO, Ada.Numerics.Long_Elementary_Functions,
Ada.Long_Float_Text_IO;
use Ada.Text_IO, Ada.Integer_Text_IO, Ada.Strings.Unbounded,
Ada.Strings.Unbounded.Text_IO, Ada.Numerics.Long_Elementary_Functions,
Ada.Long_Float_Text_IO;
procedure Fibonacci_Words is
function Entropy (S : Unbounded_String) return Long_Float is
CF : array (Character) of Natural := (others => 0);
Len : constant Natural := Length (S);
H : Long_Float := 0.0;
Ratio : Long_Float;
begin
for I in 1 .. Len loop
CF (Element (S, I)) := CF (Element (S, I)) + 1;
end loop;
for C in Character loop
Ratio := Long_Float (CF (C)) / Long_Float (Len);
if Ratio /= 0.0 then
H := H - Ratio * Log (Ratio, 2.0);
end if;
end loop;
return H;
end Entropy;
procedure Print_Line (Word : Unbounded_String; Number : Integer) is
begin
Put (Number, 4);
Put (Length (Word), 10);
Put (Entropy (Word), 2, 15, 0);
if Length (Word) < 35 then
Put (" " & Word);
end if;
New_Line;
end Print_Line;
First, Second, Result : Unbounded_String;
begin
Set_Col (4); Put ("N");
Set_Col (9); Put ("Length");
Set_Col (16); Put ("Entropy");
Set_Col (35); Put_Line ("Word");
First := To_Unbounded_String ("1");
Print_Line (First, 1);
Second := To_Unbounded_String ("0");
Print_Line (Second, 2);
for N in 3 .. 37 loop
Result := Second & First;
Print_Line (Result, N);
First := Second;
Second := Result;
end loop;
end Fibonacci_Words;

View file

@ -0,0 +1,91 @@
! Fibonacci word
! tested with Intel ifx (IFX) 2025.2.1 20250806 on Kubuntu 25.10
! GNU gfortran (Ubuntu 15.2.0-4ubuntu4) 15.2.0 on Kubuntu 25.10
! but not VSI Fortran x86-64 V8.7-001 because that compiler does not accept
! allocatable character variables
!
program FibonacciWord
character (len=:), allocatable :: fw1,fw2,fw3
integer :: i, j, ones, zeros
fw1 = '1'
fw2 = '0'
i = 2
! Print top line and first 2 lines that aren't within the do-loop
print *, 'N Length Entropy Fibword'
call printLine (1,fw1)
call printLine (2,fw2)
! Print lines 3 through 37 as requested in the task descrtiption
do j=3,37
!
! Use rotating index i instead of moving long strings around:
! calculate fw(i) as concatenation of the two other fw's in the correct order.
!
i = i+1
if (i .gt.3) i=1
select case (i)
case (1)
fw1 = fw3 // fw2
call printLine (j, fw1)
case (2)
fw2 = fw1 // fw3
call printLine (j, fw2)
case (3)
fw3 = fw2 // fw1
call printLine (j, fw3)
end select
enddo
contains
! =================================================================================================
! Print 1 output line: N, the Length of the FibWord, its Entropy, and the word unless it's too long
! =================================================================================================
subroutine printLine (ii, txt)
integer, intent(in) :: ii ! Line number
character (len=*),intent(in) :: txt ! The FibWord to print
integer :: l, i ! Length, loop index
real (kind=8) :: count_0,count_1,tot ! Count 0's and 1's in txt
real (kind=8) :: Entropy ! Resulting entropy
count_1 = 0
count_0 = 0
l = len(txt)
do i=1,l
if (txt(i:i) .eq. '0') then
count_0 = count_0 + 1_8
else
count_1 = count_1 + 1_8
endif
end do
tot = count_1+count_0
if (count_1 .eq. 0 .or. count_0 .eq. 0) then ! Calc log(0) is undefined. Use 0
Entropy =0.0_8
else
Entropy = - count_1 / tot * log2( count_1 / tot);
Entropy = Entropy - count_0 / tot * log2( count_0 / tot);
endif
if (l .lt. 60) then
write (6,'(i2,X, i8, X, F10.8, X, A )') ii, l, Entropy, txt
else
write (6,'(i2,X, i8, X, F10.8, X, A )') ii, l, Entropy, '(...)'
endif
end subroutine printLine
! =================================================
! Helper, Fortran knows LOG and LOG10 but not LOG2.
! Just a trivial base change
! =================================================
function log2 (x) result (r)
real (kind=8), intent(in) ::x
real (kind=8) ::r
r = log(x) / log(2.)
end function log2
end program FibonacciWord

View file

@ -1,9 +1,10 @@
using DataStructures
using DataStructures, Printf
entropy(s::AbstractString) = -sum(x -> x / length(s) * log2(x / length(s)), values(counter(s)))
function fibboword(n::Int64)
# Initialize the result
r = Array{String}(n)
r = fill("", n)
# First element
r[1] = "0"
# If more than 2, set the second element
@ -17,10 +18,9 @@ end
function testfibbo(n::Integer)
fib = fibboword(n)
for i in 1:length(fib)
for i in eachindex(fib)
@printf("%3d%9d%12.6f\n", i, length(fib[i]), entropy(fib[i]))
end
return 0
end
println(" n\tlength\tentropy")

View file

@ -0,0 +1,36 @@
local fmt = require "fmt"
local function entropy(s)
local m = {}
for i = 1, #s do
local c = s[i]
local d = m[c]
m[c] = (d) ? d + 1 : 1
end
local hm = 0
for m:keys() as k do
local c = m[k]
hm += c * math.log(c, 2)
end
local l = #s
return math.log(l, 2) - hm / l
end
local function fibword(n)
if n < 2 then return tostring(n) end
local a = "1"
local b = "0"
local i = 3
while i <= n do
local c = b .. a
a, b = b, c
i += 1
end
return b
end
fmt.print("%2s %10s %10s %s", "n", "Length", " Entropy ", "Fib word")
for i = 1, 37 do
local fw = fibword(i)
fmt.print("%2d %10s %0.8f %s", i, fmt.int(#fw), entropy(fw), fmt.abridge(fw, 20))
end

View file

@ -0,0 +1,76 @@
const std = @import("std");
fn findEntropy(fiboword: []const u8) f64 {
// Instead of using a hash map (like std::map in C++), using a flat
// array is extremely fast for mapping bytes to frequency counts.
var frequencies = [_]usize{0} ** 256;
for (fiboword) |c| {
frequencies[c] += 1;
}
const numlen: f64 = @floatFromInt(fiboword.len);
var infocontent: f64 = 0;
for (frequencies) |count| {
if (count > 0) {
const freq = @as(f64, @floatFromInt(count)) / numlen;
infocontent += freq * std.math.log2(freq);
}
}
// Multiply by -1 as in the original code
// Use `+ 0.0` trick to avoid negative zero (-0.0) if infocontent is 0
return -infocontent + 0.0;
}
fn printLine(writer: anytype, fiboword: []const u8, n: u32) !void {
const entropy = findEntropy(fiboword);
// Formatting equivalents:
// {d:<5} -> std::setw(5) << std::left << n
// {d:>12} -> std::setw(12) << std::right << fiboword.size()
// {d:<16.13} -> std::setw(16) << std::setprecision(13) << std::left << entropy
try writer.print("{d:<5}{d:>12} {d:<16.13}\n", .{ n, fiboword.len, entropy });
}
pub fn main() !void {
// A GeneralPurposeAllocator works perfectly for the large strings we will generate.
// Fib(37) string length expands to ~24 Megabytes.
var gpa = std.heap.GeneralPurposeAllocator(.{}){};
defer _ = gpa.deinit();
const allocator = gpa.allocator();
const stdout = std.io.getStdOut().writer();
// Print header
try stdout.print("{s:<5}{s:>12} {s:<16}\n", .{ "N", "length", "entropy" });
// We start off by allocating the base strings on the heap so they
// can be naturally freed in the memory cycle below without compiler complaints.
var first: []const u8 = try allocator.dupe(u8, "1");
var n: u32 = 1;
try printLine(stdout, first, n);
var second: []const u8 = try allocator.dupe(u8, "0");
n += 1;
try printLine(stdout, second, n);
while (n < 37) {
// Concatenate `first` and `second` into a new allocated string
const result = try std.mem.concat(allocator, u8, &.{ first, second });
// C++ uses `.assign()`, which copies the string's characters.
// Here we can simply free the oldest string, step the slice references
// forward, and save execution time avoiding memory copy overhead!
allocator.free(first);
first = second;
second = result;
n += 1;
try printLine(stdout, result, n);
}
// Clean up our remaining two string references
allocator.free(first);
allocator.free(second);
}