I need to multiplicate two matrix, but it is not the classic multiplication, instead, it is a entry by entry multiplication, as show in the next image. It is very important to know that all matrix dimensions are nxm (always), it is generic because the imput matrixs are generate by a dynamic process, the dimensions are not always same, then I need a generic and dymanic process to solve that issue.
Solved! Go to Solution.
Hey @edgarz01,
My suggestion would be to transpose the matrices, so you get the following structure:
Col Row Value
1 1 a1
1 2 a2
...
And then you do the same thing with matrix b.
When you have both matrices in long data form, you can use a join (on row and col) to combine them and then use a regular formula tool to do the multiplication. When done, transpose backwards and you should have your result.
@edgarz01
It looks to be a very interesting one.
But you are saying that it is an entry by entry multiplication, so the most right-top 2 entries in the product matrix should be
a1n * b1n
a2n * b2n?
assuming @Qiu is correct - record id. transpose (record id as key column). tile tool (record id as key column). join on record id/tile sequence number. multiple x*y. crosstab with record id as key columns. tile sequence as header. multiplication product as value.
see attached.
@apathetichell
nice work! 😁
Nice, it a simple and accurate solution. I used the suggestion and it worked like a charm.
@Qiu @FrederikE @apathetichell Thank u for your solutions, I was able to solve the problem with your suggestions.
@edgarz01
Thank you for your feedback and glad to know it is solved.