Transposing a static array
How would I transpose float[3][2] to float[2][3] with http://dlang.org/phobos/std_range.html#.transposed
Re: Transposing a static array
On 02/19/2016 06:00 PM, maik klein wrote: How would I transpose float[3][2] to float[2][3] with http://dlang.org/phobos/std_range.html#.transposed Because static arrays are not ranges, they must be used as slices with the help of []. The following code does the same thing in two different ways: import std.stdio; import std.range; import std.algorithm; void main() { float[3][2] arr = [ [1, 2, 3], [4, 5, 6] ]; // Method A { float[][] arr2; foreach (ref a; arr) { arr2 ~= a[]; } writeln(arr2.transposed); } // Method B { auto r = arr[].map!((ref a) => a[]).array.transposed; writeln(r); } } Ali
Re: Transposing a static array
On Saturday, 20 February 2016 at 02:22:12 UTC, Ali Çehreli wrote: On 02/19/2016 06:00 PM, maik klein wrote: How would I transpose float[3][2] to float[2][3] with http://dlang.org/phobos/std_range.html#.transposed Because static arrays are not ranges, they must be used as slices with the help of []. The following code does the same thing in two different ways: import std.stdio; import std.range; import std.algorithm; void main() { float[3][2] arr = [ [1, 2, 3], [4, 5, 6] ]; // Method A { float[][] arr2; foreach (ref a; arr) { arr2 ~= a[]; } writeln(arr2.transposed); } // Method B { auto r = arr[].map!((ref a) => a[]).array.transposed; writeln(r); } } Ali Your "Method B" is how I did it too but how do I convert it back to a static array of float[2][3]?
Re: Transposing a static array
On Saturday, 20 February 2016 at 02:26:56 UTC, maik klein wrote: On Saturday, 20 February 2016 at 02:22:12 UTC, Ali Çehreli wrote: On 02/19/2016 06:00 PM, maik klein wrote: How would I transpose float[3][2] to float[2][3] with http://dlang.org/phobos/std_range.html#.transposed Because static arrays are not ranges, they must be used as slices with the help of []. The following code does the same thing in two different ways: import std.stdio; import std.range; import std.algorithm; void main() { float[3][2] arr = [ [1, 2, 3], [4, 5, 6] ]; // Method A { float[][] arr2; foreach (ref a; arr) { arr2 ~= a[]; } writeln(arr2.transposed); } // Method B { auto r = arr[].map!((ref a) => a[]).array.transposed; writeln(r); } } Ali Your "Method B" is how I did it too but how do I convert it back to a static array of float[2][3]? I don't see the point of using transposed which gives a range out if you want to put it back into a float[2][3] right away, just do a double foreach: void main(string[] args) { float[3][2] src = [[1, 2, 3], [2, 3, 4]]; float[2][3] dst; foreach (i ; 0..src.length) foreach (j ; 0..src[0].length) dst[j][i] = src[i][j]; assert(dst == [[1, 2], [2, 3], [3, 4]]); }
Re: Transposing a static array
On 02/19/2016 06:26 PM, maik klein wrote: > On Saturday, 20 February 2016 at 02:22:12 UTC, Ali Çehreli wrote: >> On 02/19/2016 06:00 PM, maik klein wrote: >>> How would I transpose >>> >>> float[3][2] >>> >>> to >>> >>> float[2][3] >>> >>> with http://dlang.org/phobos/std_range.html#.transposed >> auto r = arr[].map!((ref a) => a[]).array.transposed; There is a problem there. Adding the following line writeln(r.array); causes an error: core.exception.AssertError@/usr/include/dmd/phobos/std/range/primitives.d(2207): Attempting to fetch the front of an empty array of float Adding the following two lines does *not* preserve the range (first one does): writeln(r.save); writeln(r.save); Did I not need (ref a) up there? I thought it would be a reference to a float[3] element. >> writeln(r); >> } >> } >> >> Ali > > Your "Method B" is how I did it too but how do I convert it back to a > static array of float[2][3]? I don't know. :D Ali
Re: Transposing a static array
On Saturday, 20 February 2016 at 03:02:11 UTC, cym13 wrote: On Saturday, 20 February 2016 at 02:26:56 UTC, maik klein wrote: On Saturday, 20 February 2016 at 02:22:12 UTC, Ali Çehreli wrote: [...] Your "Method B" is how I did it too but how do I convert it back to a static array of float[2][3]? I don't see the point of using transposed which gives a range out if you want to put it back into a float[2][3] right away, just do a double foreach: void main(string[] args) { float[3][2] src = [[1, 2, 3], [2, 3, 4]]; float[2][3] dst; foreach (i ; 0..src.length) foreach (j ; 0..src[0].length) dst[j][i] = src[i][j]; assert(dst == [[1, 2], [2, 3], [3, 4]]); } You are right. I restricted myself from using indices because I wanted to learn ranges but I think that working with matrices is actually much simpler by just using good old for loops.