Concat transformation¶
Description¶
The transformation Concat concatenates the input arrays into single output array.
Inputs¶
prediction.input1
— array \(A_1\) of size \(N\).prediction.input2
— array \(A_2\) of size \(M\).etc.
New inputs are added via method append()
.
Outputs¶
1) prediction.prediction
— array \(A\), filled with concatenation result \(A=\{A_1, A_2, \dotsc\}\) of size
\(N+M+\dots\)
Implementation¶
For given input arrays \(A_1\), \(A_2\), etc the output array is:
\[A = \{(A_1)_1, \dotsc, (A_1)_N,(A_2)_1, \dotsc, (A_2)_M, \dots\}\]
Any multidimensional input array is unwrapped into array using Column-major notation (like Fortran):
\[A_{ij} \rightarrow \{ A_{11}, A_{21}, \dotsc, A_{12}, A_{22}, \dotsc\},\]
in other words the slices are taken for each column.