 # mir-algorithm 3.16.0

Dlang Core Library

To use this package, run the following command in your project's root directory:

Manual usage
Put the following dependency into your project's dependences section:

## Mir Algorithm

##### Example (3 sec)
``````/+dub.sdl:
dependency "mir-algorithm" version="~>2.0.0"
+/

void main()
{
import mir.ndslice;

auto matrix = slice!double(3, 4);
matrix[] = 0;
matrix.diagonal[] = 1;

auto row = matrix;
row = 6;
assert(matrix[2, 3] == 6); // D & C index order

import mir.stdio;
matrix.writeln;
// prints [[1.0, 0.0, 0.0, 0.0], [0.0, 1.0, 0.0, 0.0], [0.0, 0.0, 1.0, 6.0]]
}
``````
##### Example (30 sec)
``````/+dub.sdl:
dependency "mir-algorithm" version="~>2.0.0"
+/
void main()
{
import mir.ndslice;
import std.stdio : writefln;

enum fmt = "%(%(%.2f %)\n%)\n";

// Magic sqaure.
// `a` is lazy, each element is computed on-demand.
auto a = magic(5).as!float;
writefln(fmt, a);

// 5x5 grid on sqaure [1, 2] x [0, 1] with values x * x + y.
// `b` is lazy, each element is computed on-demand.
auto b = linspace!float([5, 5], [1f, 2f], [0f, 1f]).map!"a * a + b";
writefln(fmt, b);

// allocate a 5 x 5 contiguous matrix
auto c = slice!float(5, 5);

c[] = transposed(a + b / 2); // no memory allocations here
// 1. b / 2 - lazy element-wise operation with scalars
// 2. a + (...) - lazy element-wise operation with other slices
// Both slices must be `contiguous` or one-dimensional.
// 3. transposed(...) - trasposes matrix view. The result is `universal` (numpy-like) matrix.
// 5. c[] = (...) -- performs element-wise assignment.
writefln(fmt, c);
}
``````

• Registered by Ilia Ki
• 3.16.0 released a year ago
• libmir/mir-algorithm
• Apache-2.0
Authors:
• Ilia Ki
• John Michael Hall
• Shigeki Karita
• Sebastian Wilzbach
• And others
• mir.date and a bit of other code is based on Phobos
Dependencies:
mir-core
Versions:
 3.21.0 2023-Aug-02 3.20.5 2023-Jul-29 3.20.4 2023-May-22 3.20.3 2023-May-11 3.20.2 2023-May-02