Highorder is a collection of useful tools for scientific computing. Most notably, it is:
- A C++ templated array class with:
- FORTRAN ordering.
- Header only compilation.
- Useful semantics.
- Automatic memory allocation for new arrays.
- Initialized to zero by default.
- Allows up to 6 dimensions, but more can be easily added.
- A Python interface:
- Cython interface. See examples/cython.
- Boost.Python interface (coming soon!)
- Still mostly a work in progress.
- MATLAB wrappers:
- Coming soon!
- include: header files
- highorder: Python library
- examples: example implementations
- cpp: C++
- cython: Cython
General C++ usage begins by including the header and creating a typedef for your data type:
#include <highorder/array.h>
typedef highorder::array<double> darray;
darray A(10,20); // Allocate a 10x20 matrix.
A(2,3) = 2.0; // Element access is with parentheses.
A(4) = 1.0; // It's up to you to use the right number of indices.
void doublify(double *x, int n) {
for (int i=0; i<n; i++) x[i] *= 2;
}
doublify(A, A.size()); // automatic casts to the data pointer.
A *= 2.0; // Double every element.
A += 3.0; // etc.
darray B(10,20);
B = 1.0; // Initialize every element of B to 1.0.
A += B; // It's up to you to ensure B has the same dimensions as A.
A *= B; // This is element wise multiplication (NOT matrix multiplication).
B = A; // Copy the data from A into the data from B.
darray C(B); // C holds a view into the data of B.
darray D(&B(0,4), 10); // D holds a view into the 5th column of B!
A.resize(10,10); // This does not reallocate the underlying memory,
// so be sure that you don't make the matrix any bigger!
A.realloc(100,100); // This frees the old memory allocated by A. Any views
// which refer to this memory are no longer valid.
// Your program may crash if you attempt to access them.
See examples/cython.