zhao.algorithmMagic.operands.matrix.ColumnDoubleMatrix Maven / Gradle / Ivy
package zhao.algorithmMagic.operands.matrix;
import zhao.algorithmMagic.core.ASDynamicLibrary;
import zhao.algorithmMagic.exception.OperatorOperationException;
import zhao.algorithmMagic.operands.RCNOperands;
import zhao.algorithmMagic.operands.vector.DoubleVector;
import zhao.algorithmMagic.utils.ASMath;
import zhao.algorithmMagic.utils.dataContainer.IntegerAndDoubles;
import java.util.*;
/**
* 带有字段名称设置的矩阵对象，在此对象中，每个维度的数值都有与之相对应的字段描述。
*
* A matrix object with field name settings. In this object, the numerical value of each dimension has its corresponding field description.
*
* @author zhao
*/
public class ColumnDoubleMatrix extends DoubleMatrix implements RCNOperands {
private final String[] Field1;
private final String[] Field2;
/**
* 构造一个具有列名属性的整形矩阵
*
* @param colNames 该矩阵中所对应的列名称
* @param rowNames 该矩阵中所对应的行名称
* @param ints 该矩阵中需要维护的数组
*/
public ColumnDoubleMatrix(String[] colNames, String[] rowNames, double[]... ints) {
super(ints);
if (ints.length > 0) {
int length = ints[0].length;
if (rowNames != null && rowNames.length > 0) {
if (rowNames.length == ints.length) {
Field2 = rowNames;
} else {
throw new OperatorOperationException("构造字段矩阵时需要注意字段名的数量与字段数据的列数一一对应！！！ERROR => RowField.length = " + rowNames.length + "\tints.rowCount = " + ints.length);
}
} else {
Field2 = new String[0];
}
if (colNames != null && colNames.length > 0) {
if (colNames.length == length) {
Field1 = colNames;
} else {
throw new OperatorOperationException("构造字段矩阵时需要注意字段名的数量与字段数据的列数一一对应！！！ERROR => ColField.length = " + colNames.length + "\tints.colCount = " + length);
}
} else {
Field1 = new String[0];
}
} else {
throw new OperatorOperationException("The array of construction matrix cannot be empty");
}
}
/**
* 构造一个矩阵，矩阵的列数量以矩阵的第一行为准！
*
* Construct a matrix, the number of columns of the matrix is based on the first row of the matrix!
*
* @param colNames 该矩阵中所对应的列名称
* @param rowNames 该矩阵中所对应的行名称
* @param ints 该矩阵中需要维护的数组
* @return matrix object
*/
public static ColumnDoubleMatrix parse(String[] colNames, String[] rowNames, double[]... ints) {
if (ints.length > 0) {
return new ColumnDoubleMatrix(colNames, rowNames, ints);
} else {
throw new OperatorOperationException("The array of construction matrix cannot be empty");
}
}
protected static void deleteRelatedFunction(double thresholdLeft, double thresholdRight, double[][] ints, double[] mid, ArrayList res, ArrayList res_f2, String[] field2) {
if (ASDynamicLibrary.isUseC()) {
for (int i = 0; i < ints.length; i++) {
double[] anInt = ints[i];
double num = ASMath.correlationCoefficient_C(anInt, mid, anInt.length);
if (num < thresholdLeft || num > thresholdRight) {
// 这个情况代表是不符合删除区间的，也就是不需要被删除的
double[] res1 = new double[anInt.length];
System.arraycopy(anInt, 0, res1, 0, anInt.length);
res_f2.add(field2[i]);
res.add(res1);
}
}
} else {
for (int i = 0; i < ints.length; i++) {
double[] anInt = ints[i];
double num = ASMath.correlationCoefficient(anInt, mid);
if (num < thresholdLeft || num > thresholdRight) {
// 这个情况代表是不符合删除区间的，也就是不需要被删除的
double[] res1 = new double[anInt.length];
System.arraycopy(anInt, 0, res1, 0, anInt.length);
res_f2.add(field2[i]);
res.add(res1);
}
}
}
}
protected static void ex(Random random1, double[][] res, String[] rowNames, int maxIndex, int i) {
int i1 = random1.nextInt(maxIndex);
double[] temp = res[i];
res[i] = res[i1];
res[i1] = temp;
String tempS = rowNames[i];
rowNames[i] = rowNames[i1];
rowNames[i1] = tempS;
}
/**
* @return 该矩阵中所对应的列名称
*/
public String[] getColFieldNames() {
return Field1.clone();
}
/**
* @return 该矩阵中对应的行名称
*/
public String[] getRowFieldNames() {
return Field2.clone();
}
@Override
public HashMap toHashMap() {
HashMap hashMap = new HashMap<>(getRowCount() + 10);
double[][] ints = toArrays();
// 开始添加数据
for (int i = 0; i < this.Field1.length; i++) {
// 将当前字段的每一个元素添加到当前字段对应的数组中
double[] tempCol = new double[ints.length];
int count = -1;
for (double[] anInt : ints) {
tempCol[++count] = anInt[i];
}
hashMap.put(this.Field1[i], DoubleVector.parse(tempCol));
}
return hashMap;
}
@Override
public String toString() {
StringBuilder stringBuilder = new StringBuilder();
{
String[] colFieldNames = this.getColFieldNames();
if (colFieldNames.length != 0) {
// 添加列字段
for (String colFieldName : colFieldNames) {
stringBuilder.append(colFieldName).append('\t');
}
if (this.Field2.length != 0) {
stringBuilder.append("rowColName\n");
} else stringBuilder.append('\n');
}
// 添加行字段与行数据
String[] rowFieldNames = this.getRowFieldNames();
double[][] doubles = this.toArrays();
if (rowFieldNames.length != 0) {
for (int i = 0; i < doubles.length; i++) {
stringBuilder
.append(Arrays.toString(doubles[i])).append('\t')
.append(rowFieldNames[i]).append('\n');
}
} else {
for (double[] doubles1 : doubles) {
stringBuilder.append(Arrays.toString(doubles1)).append("\n");
}
}
}
return "------------DoubleMatrixStart-----------\n" +
stringBuilder +
"------------DoubleMatrixEnd------------\n";
}
/**
* 去除冗余特征维度，将当前矩阵中的每一个维度都进行方差或无向差计算，并将过于稳定的冗余特征去除。
*
* Remove redundant feature dimensions, calculate variance or undirected difference of each dimension in the current matrix, and remove redundant features that are too stable.
*
* @param threshold 冗余去除阈值，代表去除的百分比，这个值应是一个小于1的数值，例如设置为0.4 代表去除掉冗余程度倒序排行中，最后40% 的维度。
*
* Redundancy removal threshold, which represents the percentage of removal, should be a value less than 1. For example, set to 0.4 to remove the last 40% of the dimensions in the reverse order of redundancy.
* @return 去除冗余特征维度之后的新矩阵
*
* New matrix after removing redundant feature dimensions
*/
@Override
public ColumnDoubleMatrix featureSelection(double threshold) {
if (threshold >= 1) throw Matrix.OPERATOR_OPERATION_EXCEPTION;
// 计算出本次要去除的维度数量
int num = (int) (getRowCount() * threshold);
if (num <= 0) {
return ColumnDoubleMatrix.parse(getColFieldNames(), getRowFieldNames(), copyToNewArrays());
} else {
// 计算出本次剩余的维度数量
num = getRowCount() - num;
// 准备好一个排序集合，存储所有的离散值结果与数组
TreeMap treeMap = new TreeMap<>(Comparator.reverseOrder());
// 将每一个维度的向量的方差计算出来
int count = -1;
for (double[] ints : this.toArrays()) {
// 计算出离散值，并将离散值与当前行编号以及当前数组添加到集合中
treeMap.put(ASMath.undirectedDifference(ints), new IntegerAndDoubles(++count, ints));
}
// 开始获取到前 num 个数组
int index = -1;
// 构建列与数据的存储控件
String[] rowNames = this.Field2.length == 0 ? null : new String[num];
double[][] res = new double[num][getColCount()];
if (rowNames == null) {
for (IntegerAndDoubles value : treeMap.values()) {
System.arraycopy(value.doubles, 0, res[++index], 0, value.doubles.length);
--num;
if (num == 0) break;
}
} else {
for (IntegerAndDoubles value : treeMap.values()) {
System.arraycopy(value.doubles, 0, res[++index], 0, value.doubles.length);
rowNames[index] = this.Field2[value.anInt];
--num;
if (num == 0) break;
}
}
return ColumnDoubleMatrix.parse(getColFieldNames(), rowNames, res);
}
}
/**
* 删除与目标索引维度相关的所有行维度，并返回新矩阵对象。
*
* Delete all row dimensions related to the target index dimension and return a new matrix object.
*
* @param index 需要被作为相关系数中心点的行编号。
*
* The row number to be used as the center point of the correlation coefficient.
* @param thresholdLeft 相关系数阈值，需要被删除的相关系数阈值区间左边界。
*
* The correlation coefficient threshold is the left boundary of the correlation coefficient threshold interval to be deleted.
* @param thresholdRight 相关系数阈值，需要被删除的相关系数阈值区间右边界。
*
* The correlation coefficient threshold is the right boundary of the correlation coefficient threshold interval to be deleted.
* @return 进行了相关维度删除之后构造出来的新矩阵
*
* The new matrix constructed after deleting relevant dimensions
*/
@Override
public ColumnDoubleMatrix deleteRelatedDimensions(int index, double thresholdLeft, double thresholdRight) {
int rowCount = getRowCount();
if (index >= 0 && index < rowCount) {
double[][] ints = toArrays();
// 获取到当前的相关系数中心序列
double[] mid = ints[index];
boolean b1 = this.Field1.length != 0;
boolean b2 = this.Field2.length != 0;
ArrayList res = new ArrayList<>(rowCount);
ArrayList res_f2 = new ArrayList<>(b2 ? this.Field2.length : 16);
// 开始进行计算
if (b1 && b2) {
deleteRelatedFunction(thresholdLeft, thresholdRight, ints, mid, res, res_f2, Field2);
return ColumnDoubleMatrix.parse(
this.Field1.clone(), res_f2.toArray(new String[0]), res.toArray(new double[0][])
);
} else if (b2) {
// 说明第 1 字段不需要添加数据
deleteRelatedFunction(thresholdLeft, thresholdRight, ints, mid, res, res_f2, Field2);
return ColumnDoubleMatrix.parse(
null, res_f2.toArray(new String[0]), res.toArray(new double[0][])
);
} else if (b1) {
// 说明第二字段不需要加数据
if (ASDynamicLibrary.isUseC()) {
for (double[] anInt : ints) {
double num = ASMath.correlationCoefficient_C(anInt, mid, anInt.length);
if (num < thresholdLeft || num > thresholdRight) {
// 这个情况代表是不符合删除区间的，也就是不需要被删除的
double[] res1 = new double[anInt.length];
System.arraycopy(anInt, 0, res1, 0, anInt.length);
res.add(res1);
}
}
} else {
for (double[] anInt : ints) {
double num = ASMath.correlationCoefficient(anInt, mid);
if (num < thresholdLeft || num > thresholdRight) {
// 这个情况代表是不符合删除区间的，也就是不需要被删除的
double[] res1 = new double[anInt.length];
System.arraycopy(anInt, 0, res1, 0, anInt.length);
res.add(res1);
}
}
}
return ColumnDoubleMatrix.parse(
this.Field1.clone(), null, res.toArray(new double[0][])
);
} else {
// 说明都不需要字段名数据
DoubleMatrix.ex(thresholdLeft, thresholdRight, ints, mid, res);
return ColumnDoubleMatrix.parse(
null, null, res.toArray(new double[0][])
);
}
} else {
return ColumnDoubleMatrix.parse(this.Field1.clone(), this.Field2.clone(), copyToNewArrays());
}
}
/**
* 获取到指定名称的行数组
*
* Get the row array with the specified name
*
* @param name 指定的行目标名称
*
* Specified row target name
* @return 一个包含当前行元素的新数组，是支持修改的。
*
* A new array containing the elements of the current row supports modification.
*/
@Override
public double[] getArrayByRowName(String name) {
int index = 0;
for (String s : this.Field2) {
if (s.equals(name)) {
return getArrayByRowIndex(index);
}
++index;
}
return new double[0];
}
/**
* 获取到指定名称的列数组
*
* Get the col array with the specified name
*
* @param name 指定的列目标名称
*
* Specified col target name
* @return 一个包含当前列元素的新数组，是支持修改的。
*
* A new array containing the elements of the current col supports modification.
*/
@Override
public double[] getArrayByColName(String name) {
int index = 0;
for (String s : this.Field1) {
if (s.equals(name)) {
return getArrayByColIndex(index);
}
++index;
}
return new double[0];
}
/**
* 将现有矩阵的转置矩阵获取到
*
* Get the transpose of an existing matrix into
*
* @return 矩阵转置之后的新矩阵
*
* new matrix after matrix transpose
*/
@Override
public ColumnDoubleMatrix transpose() {
return ColumnDoubleMatrix.parse(
this.Field2.clone(),
this.Field1.clone(),
super.transpose().toArrays()
);
}
/**
* 将本对象中的所有数据进行洗牌打乱，随机分布数据行的排列。
*
* Shuffle all the data in this object and randomly distribute the arrangement of data rows.
*
* @param seed 打乱算法中所需要的随机种子。
*
* Disrupt random seeds required in the algorithm.
* @return 打乱之后的对象。
*
* Objects after disruption.
*/
@Override
public ColumnDoubleMatrix shuffle(long seed) {
// 行是否无字段
if (this.Field2.length == 0) {
return ColumnDoubleMatrix.parse(
this.Field1.length == 0 ? null : this.Field1.clone(),
null,
ASMath.shuffle(this.copyToNewArrays(), seed, false));
} else {
// 带着行一起迭代
double[][] res = this.copyToNewArrays();
String[] rowNames = this.Field2.clone();
// 生成随机数对象
Random random = new Random();
int maxIndex = res.length - 1;
random.setSeed(seed);
for (int i = 0; i < res.length; i++) {
ex(random, res, rowNames, maxIndex, i);
}
return ColumnDoubleMatrix.parse(this.Field1.clone(), rowNames, res);
}
}
/**
* 将本对象中的所有数据进行洗牌打乱，随机分布数据行的排列。
*
* Shuffle all the data in this object and randomly distribute the arrangement of data rows.
*
* @param random1 打乱算法中所需要的随机种子。
*
* Disrupt random seeds required in the algorithm.
* @param copy 打乱时是否需要产生一个新矩阵对象，与当前对象完全脱离关系。
*
* Whether it is necessary to generate a new matrix object when disrupting, which is completely separated from the current object.
* @param length 打乱时注重的打乱次数，最终打乱会导致最多 length * 2 个元素发生位置变化。
*
* The number of disruptions that should be paid attention to when disrupting, and the final disruption will result in the position change of up to 2 elements of length *.
* @return 打乱之后的对象。
*
* Objects after disruption.
*/
public ColumnDoubleMatrix shuffle(Random random1, boolean copy, int length) {
// 行是否无字段
if (this.Field2.length == 0) {
return ColumnDoubleMatrix.parse(
this.Field1.length == 0 ? null : this.Field1.clone(),
null,
ASMath.shuffleFunction(random1, this.getRowCount(), copy ? this.copyToNewArrays() : this.toArrays(), length));
} else {
// 带着行一起迭代
double[][] res;
String[] rowNames;
if (copy) {
res = this.copyToNewArrays();
rowNames = this.Field2.clone();
} else {
res = this.toArrays();
rowNames = this.Field2;
}
// 生成随机数对象
int maxIndex = res.length - 1;
for (int i = 0; i < length; i++) {
ex(random1, res, rowNames, maxIndex, i);
}
return ColumnDoubleMatrix.parse(this.Field1.clone(), rowNames, res);
}
}
/**
* 将数据所维护的数组左移n个位置，并获取到结果数值
*
* Move the array maintained by the data to the left n positions and get the result value
*
* @param n 被左移的次数，该数值应取值于 [0, getRowCount]
*
* The number of times it is moved to the left. The value should be [0, getRowCount]
* @param copy 本次左移的作用参数，如果设置为true，代表本次位移会创建出一个新的数组，于当前数组毫无关联。
*
* If the action parameter of this left shift is set to true, it means that this shift will create a new array, which has no association with the current array.
* @return 位移之后的AS操作数对象，其类型与调用者数据类型一致。
*
* The AS operand object after displacement has the same type as the caller data type.
*/
@Override
public ColumnDoubleMatrix leftShift(int n, boolean copy) {
if (copy) {
return ColumnDoubleMatrix.parse(
this.Field1.length == 0 ? null : this.Field1.clone(),
this.Field2.length == 0 ? null : ASMath.leftShift(this.Field2.clone(), n),
ASMath.leftShift(this.copyToNewArrays(), n)
);
} else {
if (this.Field2.length != 0) {
ASMath.leftShift(this.Field2, n);
}
ASMath.leftShift(this.toArrays(), n);
return this;
}
}
/**
* 将数据所维护的数组右移n个位置，并获取到结果数值
*
* Move the array maintained by the data to the right n positions and get the result value
*
* @param n 被右移的次数，该数值应取值于 [0, getRowCount]
*
* The number of times it is moved to the right. The value should be [0, getRowCount]
* @param copy 本次右移的作用参数，如果设置为true，代表本次位移会创建出一个新的数组，于当前数组毫无关联。
*
* If the action parameter of this right shift is set to true, it means that this shift will create a new array, which has no association with the current array.
* @return 位移之后的AS操作数对象，其类型与调用者数据类型一致。
*
* The AS operand object after displacement has the same type as the caller data type.
*/
@Override
public ColumnDoubleMatrix rightShift(int n, boolean copy) {
if (copy) {
return ColumnDoubleMatrix.parse(
this.Field1.length == 0 ? null : this.Field1.clone(),
this.Field2.length == 0 ? null : ASMath.rightShift(this.Field2.clone(), n),
ASMath.rightShift(this.copyToNewArrays(), n)
);
} else {
if (this.Field2.length != 0) {
ASMath.rightShift(this.Field2, n);
}
ASMath.rightShift(this.toArrays(), n);
return this;
}
}
}