Yorick is good at manipulating elements in N-dimensional arrays conveniently with its powerful syntax.
Several elements can be accessed all at once:
Like "theading" in PDL and "broadcasting" in Numpy, Yorick has a mechanism to do this:
".." is a rubber-index to represent zero or more dimensions of the array.
"*" is a kind of rubber-index to reshape a slice(sub-array) of array to a vector.
Tensor multiplication is done as follows in Yorick:
means ∑ j = 1 j = N P i j k l Q m n j {\displaystyle \sum _{j=1}^{j=N}{P_{ijkl}Q_{mnj}}}