The rtable_scanblock(A) command applies one or more operations to the values in the specified ranges of an Array, Matrix, or Vector.

If ranges is an empty list all elements of A will be scanned.  Otherwise, the given list of ranges must be fully specified with integer lower and upper bounds, and fall within the index-space of A.  Note that for an integer $n$, the range $n..n$ can be written as simply $n$.

The following built-in operations are recognized.

Most of these built-in operations work only on numeric data.  A value of FAIL is returned if the operation cannot be completed.

The operation parameter can be a procedure that accepts either the calling sequence given by operation_passindex or by operation_noindex listed in the Calling Sequence section. If the option noindex is specified in the calling sequence, then every operation following is assumed to have the form of the operation_noindex calling sequence.  In other words, the elements index is not passed to those operation procedures.  This may be more efficient as creation of the index expression sequence can sometimes be avoided.  When passindex is specified, then the subsequent operation parameters are called with the index of the element in addition to the element's value.

The data parameter to the operation callbacks is initially the valN value passed to rtable_scanblock.  After the first callback, data becomes the result of the last callback. This is useful for building a result that depends on previous scanned elements.

It is often desirable to stop scanning once certain information is known about an element.  For example, upon detection of the first non-numeric type you may want to interrupt the scan.  This can be done using the stopresult parameter.  If the operation procedure returns the value given by stopresult, then the block scan is terminated and that result is returned.  Only the operation paired with its stopresult is terminated.  If other operations specified in the rtable_scanblock calling sequence are not finished, they will continue.

The rtable_scanblock command is thread-safe as of Maple 15.

For more information on thread safety, see index/threadsafe.

$N≔100\:$

$A≔\mathrm{LinearAlgebra}:-\mathrm{RandomMatrix}\left(N\,N\,\mathrm{outputoptions}=\left[\mathrm{datatype}=\mathrm{float}\right]\right)$

$\mathrm{rtable\_scanblock}\left(A\,\left[\mathrm{rtable\_dims}\left(A\right)\right]\,'\mathrm{Average}'\,'\mathrm{Maximum}'\,'\mathrm{Minimum}'\right)$

$0.292500000000000,99.,\mathrm{-99.}$

gzero := proc(val,res)
    if val > 0 then
        return true;
    else
        return false;
    end if;
end proc;

$\mathrm{gzero}≔\mathbf{proc}\left(\mathrm{val}\&comma;\mathrm{res}\right)\mathbf{if}0<\mathrm{val}\mathbf{then}\mathbf{return}\mathrm{true}\mathbf{else}\mathbf{return}\mathrm{false}\mathbf{end\; if}\mathbf{end\; proc}$

$A≔\mathrm{Vector}\left(\left[-1\&comma;0\&comma;1\right]\right)\&colon;$

$\mathrm{rtable\_scanblock}\left(A\&comma;\left[\right]\&comma;'\mathrm{noindex}'\&comma;\mathrm{gzero}\&comma;'\mathrm{stopresult}'=\mathrm{true}\right)$

$\mathrm{true}$

$B≔\mathrm{Vector}\left(\left[-1\&comma;0\&comma;-2\right]\right)\&colon;$

$\mathrm{rtable\_scanblock}\left(B\&comma;\left[\right]\&comma;'\mathrm{noindex}'\&comma;\mathrm{gzero}\&comma;'\mathrm{stopresult}'=\mathrm{true}\right)$

$\mathrm{false}$

$N≔100\&colon;$

$A≔\mathrm{LinearAlgebra}:-\mathrm{RandomMatrix}\left(N\&comma;N\&comma;\mathrm{outputoptions}=\left[\mathrm{datatype}=\mathrm{float}\right]\right)$

$\mathrm{rtable\_scanblock}\left(A\&comma;\left[1..N\&comma;1..1\right]\&comma;\left(\mathrm{val}\&comma;\mathrm{ind}\&comma;\mathrm{res}\right)\mapsto \mathrm{`if`}\left(\mathrm{res}\left[2\right]<\mathrm{val}\&comma;\left[\mathrm{ind}\&comma;\mathrm{val}\right]\&comma;\mathrm{res}\right)\&comma;\left[\left[1\&comma;1\right]\&comma;A\left[1,1\right]\right]\right)$

$\left[\left[94\&comma;1\right]\&comma;99.\right]$

$N≔100\&colon;$

$A≔\mathrm{LinearAlgebra}:-\mathrm{RandomMatrix}\left(N\&comma;N\&comma;\mathrm{outputoptions}=\left[\mathrm{datatype}=\mathrm{float}\right]\right)$

$\mathrm{rtable\_scanblock}\left(A\&comma;\left[N..N\&comma;1..N\right]\&comma;\left(\mathrm{val}\&comma;\mathrm{ind}\&comma;\mathrm{res}\right)\mapsto \mathrm{`if`}\left(\mathrm{res}\left[2\right]<\mathrm{val}\&comma;\left[\mathrm{ind}\&comma;\mathrm{val}\right]\&comma;\mathrm{res}\right)\&comma;\left[\left[1\&comma;1\right]\&comma;A\left[1,1\right]\right]\&comma;\left(\mathrm{val}\&comma;\mathrm{ind}\&comma;\mathrm{res}\right)\mapsto \mathrm{`if`}\left(\mathrm{val}<\mathrm{res}\left[2\right]\&comma;\left[\mathrm{ind}\&comma;\mathrm{val}\right]\&comma;\mathrm{res}\right)\&comma;\left[\left[1\&comma;1\right]\&comma;A\left[1,1\right]\right]\right)$

$\left[\left[100\&comma;4\right]\&comma;96.\right],\left[\left[100\&comma;12\right]\&comma;\mathrm{-95.}\right]$

The rtable_scanblock command was updated in Maple 2020.