/****************************************************************************
 *
 *   Copyright (C) 2012 PX4 Development Team. All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 *
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in
 *    the documentation and/or other materials provided with the
 *    distribution.
 * 3. Neither the name PX4 nor the names of its contributors may be
 *    used to endorse or promote products derived from this software
 *    without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
 * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
 * COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
 * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS
 * OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED
 * AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
 * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
 * POSSIBILITY OF SUCH DAMAGE.
 *
 ****************************************************************************/

/**
 * @file Matrix.h
 *
 * matrix code
 */

#pragma once


#include <inttypes.h>
#include <assert.h>
#include <stdlib.h>
#include <string.h>
#include <stdio.h>
#include <math.h>

#include "../Vector.hpp"
#include "../Matrix.hpp"

namespace math
{

class __EXPORT Matrix
{
public:
	// constructor
	Matrix(size_t rows, size_t cols) :
		_rows(rows),
		_cols(cols),
		_data((float *)calloc(rows *cols, sizeof(float))) {
	}
	Matrix(size_t rows, size_t cols, const float *data) :
		_rows(rows),
		_cols(cols),
		_data((float *)malloc(getSize())) {
		memcpy(getData(), data, getSize());
	}
	// deconstructor
	virtual ~Matrix() {
		delete [] getData();
	}
	// copy constructor (deep)
	Matrix(const Matrix &right) :
		_rows(right.getRows()),
		_cols(right.getCols()),
		_data((float *)malloc(getSize())) {
		memcpy(getData(), right.getData(),
		       right.getSize());
	}
	// assignment
	inline Matrix &operator=(const Matrix &right) {
#ifdef MATRIX_ASSERT
		ASSERT(getRows() == right.getRows());
		ASSERT(getCols() == right.getCols());
#endif

		if (this != &right) {
			memcpy(getData(), right.getData(),
			       right.getSize());
		}

		return *this;
	}
	// element accessors
	inline float &operator()(size_t i, size_t j) {
#ifdef MATRIX_ASSERT
		ASSERT(i < getRows());
		ASSERT(j < getCols());
#endif
		return getData()[i * getCols() + j];
	}
	inline const float &operator()(size_t i, size_t j) const {
#ifdef MATRIX_ASSERT
		ASSERT(i < getRows());
		ASSERT(j < getCols());
#endif
		return getData()[i * getCols() + j];
	}
	// output
	inline void print() const {
		for (size_t i = 0; i < getRows(); i++) {
			for (size_t j = 0; j < getCols(); j++) {
				float sig;
				int exp;
				float num = (*this)(i, j);
				float2SigExp(num, sig, exp);
				printf("%6.3fe%03.3d,", (double)sig, exp);
			}

			printf("\n");
		}
	}
	// boolean ops
	inline bool operator==(const Matrix &right) const {
		for (size_t i = 0; i < getRows(); i++) {
			for (size_t j = 0; j < getCols(); j++) {
				if (fabsf((*this)(i, j) - right(i, j)) > 1e-30f)
					return false;
			}
		}

		return true;
	}
	// scalar ops
	inline Matrix operator+(const float &right) const {
		Matrix result(getRows(), getCols());

		for (size_t i = 0; i < getRows(); i++) {
			for (size_t j = 0; j < getCols(); j++) {
				result(i, j) = (*this)(i, j) + right;
			}
		}

		return result;
	}
	inline Matrix operator-(const float &right) const {
		Matrix result(getRows(), getCols());

		for (size_t i = 0; i < getRows(); i++) {
			for (size_t j = 0; j < getCols(); j++) {
				result(i, j) = (*this)(i, j) - right;
			}
		}

		return result;
	}
	inline Matrix operator*(const float &right) const {
		Matrix result(getRows(), getCols());

		for (size_t i = 0; i < getRows(); i++) {
			for (size_t j = 0; j < getCols(); j++) {
				result(i, j) = (*this)(i, j) * right;
			}
		}

		return result;
	}
	inline Matrix operator/(const float &right) const {
		Matrix result(getRows(), getCols());

		for (size_t i = 0; i < getRows(); i++) {
			for (size_t j = 0; j < getCols(); j++) {
				result(i, j) = (*this)(i, j) / right;
			}
		}

		return result;
	}
	// vector ops
	inline Vector operator*(const Vector &right) const {
#ifdef MATRIX_ASSERT
		ASSERT(getCols() == right.getRows());
#endif
		Vector result(getRows());

		for (size_t i = 0; i < getRows(); i++) {
			for (size_t j = 0; j < getCols(); j++) {
				result(i) += (*this)(i, j) * right(j);
			}
		}

		return result;
	}
	// matrix ops
	inline Matrix operator+(const Matrix &right) const {
#ifdef MATRIX_ASSERT
		ASSERT(getRows() == right.getRows());
		ASSERT(getCols() == right.getCols());
#endif
		Matrix result(getRows(), getCols());

		for (size_t i = 0; i < getRows(); i++) {
			for (size_t j = 0; j < getCols(); j++) {
				result(i, j) = (*this)(i, j) + right(i, j);
			}
		}

		return result;
	}
	inline Matrix operator-(const Matrix &right) const {
#ifdef MATRIX_ASSERT
		ASSERT(getRows() == right.getRows());
		ASSERT(getCols() == right.getCols());
#endif
		Matrix result(getRows(), getCols());

		for (size_t i = 0; i < getRows(); i++) {
			for (size_t j = 0; j < getCols(); j++) {
				result(i, j) = (*this)(i, j) - right(i, j);
			}
		}

		return result;
	}
	inline Matrix operator*(const Matrix &right) const {
#ifdef MATRIX_ASSERT
		ASSERT(getCols() == right.getRows());
#endif
		Matrix result(getRows(), right.getCols());

		for (size_t i = 0; i < getRows(); i++) {
			for (size_t j = 0; j < right.getCols(); j++) {
				for (size_t k = 0; k < right.getRows(); k++) {
					result(i, j) += (*this)(i, k) * right(k, j);
				}
			}
		}

		return result;
	}
	inline Matrix operator/(const Matrix &right) const {
#ifdef MATRIX_ASSERT
		ASSERT(right.getRows() == right.getCols());
		ASSERT(getCols() == right.getCols());
#endif
		return (*this) * right.inverse();
	}
	// other functions
	inline Matrix transpose() const {
		Matrix result(getCols(), getRows());

		for (size_t i = 0; i < getRows(); i++) {
			for (size_t j = 0; j < getCols(); j++) {
				result(j, i) = (*this)(i, j);
			}
		}

		return result;
	}
	inline void swapRows(size_t a, size_t b) {
		if (a == b) return;

		for (size_t j = 0; j < getCols(); j++) {
			float tmp = (*this)(a, j);
			(*this)(a, j) = (*this)(b, j);
			(*this)(b, j) = tmp;
		}
	}
	inline void swapCols(size_t a, size_t b) {
		if (a == b) return;

		for (size_t i = 0; i < getRows(); i++) {
			float tmp = (*this)(i, a);
			(*this)(i, a) = (*this)(i, b);
			(*this)(i, b) = tmp;
		}
	}
	/**
	 * inverse based on LU factorization with partial pivotting
	 */
	Matrix inverse() const {
#ifdef MATRIX_ASSERT
		ASSERT(getRows() == getCols());
#endif
		size_t N = getRows();
		Matrix L = identity(N);
		const Matrix &A = (*this);
		Matrix U = A;
		Matrix P = identity(N);

		//printf("A:\n"); A.print();

		// for all diagonal elements
		for (size_t n = 0; n < N; n++) {

			// if diagonal is zero, swap with row below
			if (fabsf(U(n, n)) < 1e-8f) {
				//printf("trying pivot for row %d\n",n);
				for (size_t i = 0; i < N; i++) {
					if (i == n) continue;

					//printf("\ttrying row %d\n",i);
					if (fabsf(U(i, n)) > 1e-8f) {
						//printf("swapped %d\n",i);
						U.swapRows(i, n);
						P.swapRows(i, n);
					}
				}
			}

#ifdef MATRIX_ASSERT
			//printf("A:\n"); A.print();
			//printf("U:\n"); U.print();
			//printf("P:\n"); P.print();
			//fflush(stdout);
			ASSERT(fabsf(U(n, n)) > 1e-8f);
#endif

			// failsafe, return zero matrix
			if (fabsf(U(n, n)) < 1e-8f) {
				return Matrix::zero(n);
			}

			// for all rows below diagonal
			for (size_t i = (n + 1); i < N; i++) {
				L(i, n) = U(i, n) / U(n, n);

				// add i-th row and n-th row
				// multiplied by: -a(i,n)/a(n,n)
				for (size_t k = n; k < N; k++) {
					U(i, k) -= L(i, n) * U(n, k);
				}
			}
		}

		//printf("L:\n"); L.print();
		//printf("U:\n"); U.print();

		// solve LY=P*I for Y by forward subst
		Matrix Y = P;

		// for all columns of Y
		for (size_t c = 0; c < N; c++) {
			// for all rows of L
			for (size_t i = 0; i < N; i++) {
				// for all columns of L
				for (size_t j = 0; j < i; j++) {
					// for all existing y
					// subtract the component they
					// contribute to the solution
					Y(i, c) -= L(i, j) * Y(j, c);
				}

				// divide by the factor
				// on current
				// term to be solved
				// Y(i,c) /= L(i,i);
				// but L(i,i) = 1.0
			}
		}

		//printf("Y:\n"); Y.print();

		// solve Ux=y for x by back subst
		Matrix X = Y;

		// for all columns of X
		for (size_t c = 0; c < N; c++) {
			// for all rows of U
			for (size_t k = 0; k < N; k++) {
				// have to go in reverse order
				size_t i = N - 1 - k;

				// for all columns of U
				for (size_t j = i + 1; j < N; j++) {
					// for all existing x
					// subtract the component they
					// contribute to the solution
					X(i, c) -= U(i, j) * X(j, c);
				}

				// divide by the factor
				// on current
				// term to be solved
				X(i, c) /= U(i, i);
			}
		}

		//printf("X:\n"); X.print();
		return X;
	}
	inline void setAll(const float &val) {
		for (size_t i = 0; i < getRows(); i++) {
			for (size_t j = 0; j < getCols(); j++) {
				(*this)(i, j) = val;
			}
		}
	}
	inline void set(const float *data) {
		memcpy(getData(), data, getSize());
	}
	inline size_t getRows() const { return _rows; }
	inline size_t getCols() const { return _cols; }
	inline static Matrix identity(size_t size) {
		Matrix result(size, size);

		for (size_t i = 0; i < size; i++) {
			result(i, i) = 1.0f;
		}

		return result;
	}
	inline static Matrix zero(size_t size) {
		Matrix result(size, size);
		result.setAll(0.0f);
		return result;
	}
	inline static Matrix zero(size_t m, size_t n) {
		Matrix result(m, n);
		result.setAll(0.0f);
		return result;
	}
protected:
	inline size_t getSize() const { return sizeof(float) * getRows() * getCols(); }
	inline float *getData() { return _data; }
	inline const float *getData() const { return _data; }
	inline void setData(float *data) { _data = data; }
private:
	size_t _rows;
	size_t _cols;
	float *_data;
};

} // namespace math