Create a two-dimensional array at runtime
You are encouraged to solve this task according to the task description, using any language you may know.
Data Structure
This illustrates a data structure, a means of storing data within a program.
Get two integers from the user, then create a two-dimensional array where the two dimensions have the sizes given by those numbers, and which can be accessed in the most natural way possible. Write some element of that array, and then output that element. Finally destroy the array if not done by the language itself.
Ada
<lang ada>
with Ada.Text_Io; use Ada.Text_Io;
with Ada.Float_Text_Io; use Ada.Float_Text_Io;
with Ada.Integer_Text_Io; use Ada.Integer_Text_Io;
procedure Two_Dimensional_Arrays is
type Matrix_Type is array(Positive range <>, Positive range <>) of Float;
Dim_1 : Positive;
Dim_2 : Positive;
begin
Get(Item => Dim_1);
Get(Item => Dim_2);
-- Create an inner block with the correctly sized array
declare
Matrix : Matrix_Type(1..Dim_1, 1..Dim_2);
begin
Matrix(1, Dim_2) := 3.14159;
Put(Item => Matrix(1, Dim_2), Fore => 1, Aft => 5, Exp => 0);
New_Line;
end;
-- The variable Matrix is popped off the stack automatically
end Two_Dimensional_Arrays;
</lang>
ALGOL 68
main:(
print("Input two positive whole numbers separated by space and press newline:");
[read int,read int] INT array;
array[1,1]:=42;
print (array[1,1])
)
AWK
AWK has no multidimensional array; but AWK arrays (which are Associative array indeed) can be used also in a multidimensional fashion. Since AWK arrays are associative arrays, there's no issue in their dimensions: they grow while adding new key-value pair.
<lang awk>/[0-9]+ [0-9]+/ {
for(i=0; i < $1; i++) {
for(j=0; j < $2; j++) {
arr[i, j] = i*j
}
}
# how to scan "multidim" array as explained in the GNU AWK manual
for (comb in arr) {
split(comb, idx, SUBSEP)
print idx[1] "," idx[2] "->" arr[idx[1], idx[2]]
}
}</lang>
BASIC
CLS INPUT a, b 'inputs need to be separated by commas DIM array (1 TO a, 1 TO b) array(1,1) = 42 PRINT array(1,1) ERASE array
C
<lang c>
- include <stdio.h>
int main(int argc, char **argv) {
int user1 = 0, user2 = 0;
printf("Enter two integers. Space delimited, please: ");
scanf("%d %d",&user1, &user2);
int array[user1][user2];
array[user1/2][user2/2] = user1 + user2;
printf("array[%d][%d] is %d\n",user1/2,user2/2,array[user1/2][user2/2]);
return 0;
} </lang>
C++
With language built-in facilities:
<lang cpp> #include <iostream>
#include <istream>
#include <ostream>
int main()
{
// read values
int dim1, dim2;
std::cin >> dim1 >> dim2;
// create array
double* array_data = new double[dim1*dim2];
double** array = new double*[dim1];
for (int i = 0; i < dim1; ++i)
array[i] = array_data + dim2*i;
// write element
array[0][0] = 3.5;
// output element
std::cout << array[0][0] << std::endl;
// get rid of array
delete[] array;
delete[] array_data;
}</lang>
Using std::vector from the standard library:
<lang cpp> #include <iostream>
#include <istream>
#include <ostream>
#include <vector>
int main()
{
// read values
int dim1, dim2;
std::cin >> dim1 >> dim2;
// create array
std::vector<std::vector<double> > array(dim1, std::vector<double>(dim2));
// write element
array[0][0] = 3.5;
// output element
std::cout << array[0][0] << std::endl;
// the array is automatically freed at the end of main()
}</lang>
D
<lang d> import std.stdio: writef, writefln, readln; import std.conv: toInt; import std.string: strip;
void main() {
writef("Give me the numer of rows: ");
int nrow = toInt(readln().strip());
writef("Give me the numer of columns: ");
int ncol = toInt(readln().strip());
auto array = new float[][](nrow, ncol);
array[0][0] = 3.5;
writefln("The number at place [0 0] is ", array[0][0]);
} </lang>
Clean
<lang clean> import StdEnv
Start :: *World -> { {Real} }
Start world
# (console, world) = stdio world
(_, dim1, console) = freadi console
(_, dim2, console) = freadi console
= createArray dim1 (createArray dim2 1.0)</lang>
Common Lisp
<lang lisp> (let ((d1 (read))
(d2 (read)))
(assert (and (typep d1 '(integer 1))
(typep d2 '(integer 1)))
(d1 d2))
(let ((array (make-array (list d1 d2) :initial-element nil))
(p1 0)
(p2 (floor d2 2)))
(setf (aref array p1 p2) t)
(print (aref array p1 p2))))</lang>
The assert will allow the user to reenter the dimensions if they are not positive integers.
Forth
<lang forth> : cell-matrix
create ( width height "name" ) over , * cells allot does> ( x y -- addr ) dup cell+ >r @ * + cells r> + ; 5 5 cell-matrix test 36 0 0 test ! 0 0 test @ . \ 36</lang>
<lang forth>
INTEGER DMATRIX my-matrix{{
& my-matrixTemplate:8 9malloc
8 my-matrixTemplate:3 4 !
my-matrixTemplate:3 4 @ .
& my-matrix{{ }}free
</lang>
Fortran
In Fortran 90 and later <lang fortran> PROGRAM Example
IMPLICIT NONE INTEGER :: rows, columns, errcheck INTEGER, ALLOCATABLE :: array(:,:) WRITE(*,*) "Enter number of rows" READ(*,*) rows WRITE(*,*) "Enter number of columns" READ(*,*) columns ALLOCATE (array(rows,columns), STAT=errcheck) ! STAT is optional and is used for error checking array(1,1) = 42 WRITE(*,*) array(1,1) DEALLOCATE (array, STAT=errcheck) END PROGRAM Example</lang>
Haskell
<lang haskell> doit n m = a!(0,0) where a = array ((0,0),(n,m)) [((0,0),42)]</lang>
IDL
The following is only for demonstration. No real program should just assume that the user input is valid, integer, large enough etc.
<lang idl> read, x, prompt='Enter x size:'
read, y, prompt='Enter y size:' d = fltarr(x,y) d[3,4] = 5.6 print,d[3,4] ;==> outputs 5.6 delvar, d</lang>
J
In J, all aspects of arrays are resolved through evaluation. Everything counts as being given at run time. <lang j> task=: 3 : 0
'init new' =. 0;1 NB. values for initialization and alteration
array =. y $ init NB. create array of shape y
element =. < ? $ array NB. pick an atom of the array at random
array =. new element } array NB. amend that element to new value
element { array NB. return value of changed element
)</lang>
Passing two integers to task (as a list) satisfies the specifications for a two-dimensional array, but providing a longer list of integers accomplishes the same task on an array of as many dimensions as the count of integers given.
The type of the array is determined by the type of the values used in filling the array. E.g., alternate data types are obtained by substituting any of the following lines: <lang j> 'init new' =. ' ';'x' NB. literals
'init new' =. 1r2;2r3 NB. fractions 'init new' =. a: ; <<'Rosetta' NB. boxes</lang>
Java
<lang java> import java.util.Scanner;
public class twoDimArray {
public static void main(String[] args) {
Scanner in = new Scanner(System.in);
int nbr1 = in.nextInt();
int nbr2 = in.nextInt();
double[][] array = new double[nbr1][nbr2];
array[0][0] = 42.0;
System.out.println("The number at place [0 0] is " + array[0][0]);
}
}</lang>
Logo
<lang logo> make "a2 mdarray [5 5]
mdsetitem [1 1] :a2 0 ; by default, arrays are indexed starting at 1 print mditem [1 1] :a2 ; 0</lang>
MAXScript
a = getKBValue prompt:"Enter first dimension:"
b = getKBValue prompt:"Enter second dimension:"
arr1 = #()
arr2 = #()
arr2[b] = undefined
for i in 1 to a do
(
append arr1 (deepCopy arr2)
)
arr1[a][b] = 1
print arr1[a][b]
Objective-C
Being Objective-C derivated from C, the C solution works fine in Objective-C too.
The "OpenStep" frameworks (GNUstep, Cocoa) does not provide a class for multidimensional array; of course it can be implemented in several way (also as a wrapper for the plain C way of handling arrays). Here I show a straightforward use of the NSMutableArray class.
<lang objc>#import <Foundation/Foundation.h>
int main() {
int num1, num2, i, j; NSMutableArray *arr;
NSAutoreleasePool *pool = [[NSAutoreleasePool alloc] init];
scanf("%d %d", &num1, &num2);
NSLog(@"%d %d", num1, num2); arr = [NSMutableArray arrayWithCapacity: (num1*num2)]; // initialize it with 0s for(i=0; i < (num1*num2); i++) [arr addObject: [NSNumber numberWithInt: 0]];
// replace 0s with something more interesting
for(i=0; i < num1; i++) {
for(j=0; j < num2; j++) {
[arr replaceObjectAtIndex: (j*num1+i) withObject: [NSNumber numberWithInt: (i*j)]];
}
}
// access a value: I*num1+J, where I,J are the indexes for the bidimensional array NSLog(@"%@", [arr objectAtIndex: (1*num1+3)]); [pool release]; return 0;
}</lang>
OCaml
<lang ocaml> let nbr1 = read_int ();;
let nbr2 = read_int ();; let array = Array.make_matrix nbr1 nbr2 0.0;; array.(0).(0) <- 3.5;; print_float array.(0).(0); print_newline ();;</lang>
Pascal
The following code is standard Extended Pascal (tested with gpc --extended-pascal):
<lang pascal> program array2d(input, output);
type
tArray2d(dim1, dim2: integer) = array[1 .. dim1, 1 .. dim2] of real;
pArray2D = ^tArray2D;
var
d1, d2: integer;
data: pArray2D;
begin
{ read values }
readln(d1, d2);
{ create array }
new(data, d1, d2);
{ write element }
data^[1,1] := 3.5;
{ output element }
writeln(data^[1,1]);
{ get rid of array }
dispose(data);
end.</lang>
Perl
Predefining an array (or multi-dimension array) size is unnecessary, Perl dynamically resizes the array to meet the requirements. Of course I'm assuming that the user is entering array size 0 based.
<lang perl> sub make_array($ $){
# get array sizes from provided params, but force numeric value
my $x = ($_[0] =~ /^\d+$/) ? shift : 0;
my $y = ($_[0] =~ /^\d+$/) ? shift : 0;
# define array, then add multi-dimensional elements
my @array;
$array[0][0] = 'X '; # first by first element
$array[5][7] = 'X ' if (5 <= $y and 7 <= $x); # sixth by eighth element, if the max size is big enough
$array[12][15] = 'X ' if (12 <= $y and 15 <= $x); # thirteenth by sixteenth element, if the max size is big enough
# loop through the elements expected to exist base on input, and display the elements contents in a grid
foreach my $dy (0 .. $y){
foreach my $dx (0 .. $x){
(defined $array[$dy][$dx]) ? (print $array[$dy][$dx]) : (print '. ');
}
print "\n";
}
}</lang>
Pop11
<lang pop11> vars itemrep;
incharitem(charin) -> itemrep; ;;; Read sizes vars n1 = itemrep(), n2= itemrep(); ;;; Create 0 based array vars ar = newarray([0 ^(n1 - 1) 0 ^(n2 - 1)], 0); ;;; Set element value 15 -> ar(0, 0); ;;; Print element value ar(0,0) => ;;; Make sure array is unreferenced 0 -> ar;</lang>
Pop11 is garbage collected so there is no need to destroy array. However, the array is live as long as variable ar references it. The last assignment makes sure that we loose all our references to the array turning it into garbage.
Pop11 arrays may have arbitrary lower bounds, since we are given only size we create 0 based array.
Python
<lang python>width = int(raw_input("Width of myarray: ")) height = int(raw_input("Height of Array: ")) myarray = [[0] * width for i in xrange(height)] myarray[0][0] = 3.5 print myarray[0][0]</lang>
Note: Some people may instinctively try to write myarray as [[0] * width] * height, but the * operator creates n references to [[0] * width]
You can also use a two element tuple to index a dictionary like so:
<lang python>myarray = dict(((w,h), 0) for w in range(width) for h in range(height))
- or, in Python 3: myarray = {(w,h): 0 for w in range(width) for h in range(height)}
myarray[(0,0)] = 3.5 print myarray[(0,0)]</lang>
Ruby
<lang ruby> puts 'Enter width and height: '
w=gets.to_i
arr = Array.new(gets.to_i){Array.new(w)}
arr[1][3] = 5
p arr[1][3]</lang>
Smalltalk
Smalltalk has no problems in creating objects at runtime. I haven't found a class for multidimensional array in the standard library, so let us suppose to have a class named MultidimensionalArray.
<lang smalltalk>|num1 num2 arr| num1 := stdin nextLine asInteger. num2 := stdin nextLine asInteger.
arr := MultidimensionalArray new: { num1. num2 }.
1 to: num1 do: [ :i |
1 to: num2 do: [ :j |
arr at: { i. j } put: (i*j)
]
].
1 to: num1 do: [ :i |
1 to: num2 do: [ :j |
(arr at: {i. j}) displayNl
]
].</lang>
A possible implementation for a BidimensionalArray class is the following (changing Multi into Bi and using this class, the previous code runs fine):
<lang smalltalk>Object subclass: BidimensionalArray [
|biArr| <comment: 'bidim array'>
]. BidimensionalArray class extend [
new: biDim [ |r| r := super new. r init: biDim. ^ r ]
]. BidimensionalArray extend [
init: biDim [
biArr := Array new: (biDim at: 1).
1 to: (biDim at: 1) do: [ :i |
biArr at: i put: (Array new: (biDim at: 2))
].
^ self
]
at: biDim [
^ (biArr at: (biDim at: 1)) at: (biDim at: 2)
]
at: biDim put: val [
^ (biArr at: (biDim at: 1)) at: (biDim at: 2) put: val
]
].</lang>
Instead of implementing such a class (or the MultidimensionalArray one), we can use a LookupTable class, using Array objects as keys (each element of the array will be an index for a specific dimension of the "array"). The final effect is the same as using an array (almost in the AWK sense) and the approach has some advantages.
<lang smalltalk>|num1 num2 pseudoArr| num1 := stdin nextLine asInteger. num2 := stdin nextLine asInteger.
"we can 'suggest' an initial value for the number
of slot the table can hold; anyway, if we use more than these, the table automatically grows"
pseudoArr := LookupTable new: (num1 * num2).
1 to: num1 do: [ :i |
1 to: num2 do: [ :j |
pseudoArr at: {i. j} put: (i * j).
]
].
1 to: num1 do: [ :i |
1 to: num2 do: [ :j |
(pseudoArr at: {i. j}) displayNl.
]
].</lang>
Tcl
<lang tcl>package require Tcl 8.5
puts "enter dimension:" set dim1 [gets stdin] puts "enter dimension:" set dim2 [gets stdin] set l [lrepeat $dim1 [lrepeat $dim2 {}]] lset l end end aValue puts [lindex $l end end] unset l</lang>
Toka
Toka has no direct support for 2D arrays, but they can be created and operated on in a manner similar to normal arrays using the following functions.
<lang toka> [ ( x y -- address )
cells malloc >r dup cells >r [ r> r> r> 2dup >r >r swap malloc swap i swap array.put >r ] iterate r> r> nip ] is 2D-array [ ( a b address -- value ) array.get array.get ] is 2D-get-element [ ( value a b address -- ) array.get array.put ] is 2D-put-element</lang>
And a short test:
<lang toka> 5 5 2D-array >r #! Create an array and save the pointer to it
10 2 3 r@ 2D-put-element #! Set element 2,3 to 10 2 3 r@ 2D-get-element #! Get the element at 2,3 r> drop #! Discard the pointer to the array</lang>