is_local_max.hhs

/**
 * @author Jianan Lin (林家南)
 * @param input - an array, matrix, tensor that you want to get local max
 * @param dim - 0 or 1: 0 is local max of each row, 1 is each column. 
 * dim only supports 2-d matrix. For tensor or array, this para is ignored.
 * @returns - a matrix / tensor with elements 0 or 1 denoting whether it is local max.
 * 
 */


function is_local_max(input, dim = 0) {

    *import math: deep_copy
    *import math: ndim

    // argument check
    if (arguments.length === 0) {
        throw new Error('Exception occurred in is_local_max - no argument given');
    }

    if (arguments.length > 2) {
        throw new Error('Exception occurred in is_local_max - wrong argument number');
    }

    if (!(Array.isArray(input)) && !(input instanceof Mat) && !(input instanceof Tensor)) {
        throw new Error('Exception occurred in is_local_max - input must be an array, matrix or tensor');
    }

    if (dim !== 0 && dim !== 1) {
        throw new Error('Exception occurred in is_local_max - dim must be 0 or 1');
    }

    let in_type = input instanceof Mat || input instanceof Tensor;
    let raw_in = in_type ? input.clone().val : deep_copy(input);
    let result = deep_copy(raw_in);

    // row local max
    if (dim === 0) {
        is_local_max_helper_row(raw_in, result);
        if (ndim(result) <= 2) {
            return mat(result);
        }
        else {
            return new Tensor(result);
        }
    }

    // column local max, require m * n matrix
    else {
        if (ndim(raw_in) !== 2) {
            throw new Error('Exception occurred in is_local_max - if dim = 1, then input must be a m * n matrix');
        }
        else if (is_normal_matrix(raw_in) === false) {
            throw new Error('Exception occurred in is_local_max - if dim = 1, then input must be a m * n matrix');
        }

        else {
            for (let i = 0; i < raw_in.length; i++) {
                for (let j = 0; j < raw_in[i].length; j++) {
                    if (i === 0 || i === raw_in.length - 1) {
                        result[i][j] = 0;
                    }
                    else if (raw_in[i][j] < raw_in[i - 1][j] || raw_in[i][j] < raw_in[i + 1][j]) {
                        result[i][j] = 0;
                    }
                    else {
                        result[i][j] = 1;
                    }
                }
            }

            return mat(result);
        }
    }

    function is_normal_matrix(input) {
        let m = input.length;
        let n = input[0].length;
        for (let i = 0; i < m; i++) {
            if (input[i].length !== n) {
                return false;
            }
        }
        return true;
    }

    function judge_row(array, index) {
        if (index === 0 || index === array.length - 1) {
            return 0;
        }
        else if (array[index] < array[index + 1] || array[index] < array[index - 1]) {
            return 0;
        }
        else {
            return 1;
        }
    }

    function is_local_max_helper_row(array1, array2) {
        if (ndim(array1) === 1) {
            for (let i = 0; i < array1.length; i++) {
                array2[i] = judge_row(array1, i);
            }
        }
        else {
            for (let i = 0; i < array1.length; i++) {
                is_local_max_helper_row(array1[i], array2[i]);
            }
        }
    }
}