The Insert command inserts 1-D rtable or list B into 1-D rtable or list A at index k, and returns the combined container.

When A is an rtable, the rtable order and subtype of the output are the same as for A. When A is a list, on the other hand, the output is an Array having Fortran order.

Any passed rtables must have no indexing function, and use rectangular storage.

Suppose $m=\mathrm{numelems}\left(A\right)$, $n=\mathrm{numelems}\left(B\right)$, and $A$ has dimensions $\mathrm{\alpha }..\mathrm{\beta }$. The result of $\mathrm{Insert}\left(A\&comma;k\&comma;B\right)$ contains $m+n$ elements, with $B$ inserted to the left of element $A_k$. When $A$ is an Array with initial index different than $1$, the insertion index $k$ must satisfy $\mathrm{\alpha }\le k$ and $k\le \mathrm{\beta }+1$. When $A$ is a Vector or an Array with initial index $1$, on the other hand, the insertion index must satisfy $k\ne 0$ and $\left|k\right|\le m+1$, with $k<0$ corresponding to element $m+1-k$.

The extreme case $k=\mathrm{\alpha }$ corresponds to inserting $B$ before $A$, and $k=\mathrm{\beta }+1$ corresponds to inserting $B$ after $A$.

Internally, A and B will both be converted to float[8] Vectors, if possible. If this is not possible, they will both be converted to complex[8] Vectors, if possible. For this reason, it is more efficient for the passed containers A and B to both be 1-D rtables of the same double precision hardware float datatype (float[8] or complex[8]).

Suppose inplace=true. The container A must have either float[8] or complex[8] datatype, and the container B must be coercible to the same datatype. Moreover, if B is an alias of A, or A and B are both aliases of the same rtable, then a copy of B is created in order to avoid complications during insertion.

The Insert command is not thread safe.

$\mathrm{with}\left(\mathrm{SignalProcessing}\right)\&colon;$

The SignalProcessing[Insert] command was introduced in Maple 2022.

For more information on Maple 2022 changes, see Updates in Maple 2022.