Row-Based Functions

For most DolphinDB functions that are applicable to matrices or tables, the calculation is usually conducted based on columns. To calculate on rows, you can use row-based functions.

Introduction

Higher-order function byRow:
```
byRow(func, X)
```
Syntax template for unary row-based functions:
```
rowFunc(args…)
```
Additional unary function template (released with version 2.00.10)
```
rowFunc(X)
```
Parameters:

args can be a scalar/matrix, or one or more vectors/tuples/tables or their combination. The length of vector, the number of elements of a tuple, and the number of table columns must be the same.

Note: If args is a matrix, return a vector of the same length as the number of rows of the matrix. Please note that multiple matrices are not supported in row functions.

X is a matrix, array vector, columnar tuple, or tuple with equal-length vectors. Note that the following functions do not support tuple with equal-length vectors: rowRank, rowDenseRank, rowMove, rowNext, rowPrev.
Syntax template for binary row-based functions:
```
rowFunc(X, Y)
```
Parameters:

X and Y are matrices/numeric vectors/array vectors of the same size.

The following binary functions accept two columnar tuples as arguments: rowWavg, rowCorr, rowCovar, rowBeta, and rowWsum, rowCumwsum.

List of Functions

rowFunc(args…)
rowAnd, rowOr, rowXor, rowMax, rowMin, rowCount, rowSize, rowSum, rowSum2, rowAvg, rowProd, rowStd, rowStdp, rowVar, rowVarp
rowFunc(X, Y)
rowAt, rowBeta, rowCorr, rowCovar, rowWavg, rowWsum
Other Row-Based Functions:
(1) args is a matrix:
rowRank, rowDenseRank
(2) args is a tuple, matrix, or table
rowKurtosis, rowSkew

Calculation Rules

To calculate on different data forms, please see the following example.

vec = [1,2,3]
vec_tuple = [[3,4,5],[4,5,6]]
tb = table(7 8 9 as id, 8 9 10 as code)
print rowSum(vec, vec_tuple, tb)
// output
[23, 28, 33]

To calculation on a matrix:

m = matrix(4 2 1 3 5 8, 1 2 5 9 0 1, 3 6 3 2 1 5)
print rowSum(m)
// output
[8, 10, 9, 14, 6, 14]