2

Is it possible to feed data from an org table into calc as a matrix, and then get it out again as a vector to fill in a column?

For example, if I have:

| | Fund A | Fund B | Fund C | Combined | | US | .1 | .8 | .5 | | | Europe | .2 | .1 | .4 | | | Pacific | .7 | .1 | .1 | | | Weighting | .3 | .5 | .2 | |

I would like to use a calc formula to matrix-multiply the data from the first three rows of numbers by the (transposed) weighting in the last row of numbers, and insert the result in the final column, like this:

| | Fund A | Fund B | Fund C | Combined | | US | .1 | .8 | .5 | .53 | | Europe | .2 | .1 | .4 | .19 | | Pacific | .7 | .1 | .1 | .28 | | Weighting | .3 | .5 | .2 | |

I was hoping maybe something like this would achieve that if inserted in the top cell of the last column -- but this is incorrect: no output is produced in the table:

=($2@2..$2@4) * trn($4@2..$4..@4)

If it's not possible using calc, is it possible using elisp?

Improve this question

1 Answer 1

Reset to default
3

The following works for me:

| | Fund A | Fund B | Fund C | Combined | |-----------+--------+--------+--------+----------| | US | .1 | .8 | .5 | 0.53 | | Europe | .2 | .1 | .4 | 0.19 | | Pacific | .7 | .1 | .1 | 0.28 | | Weighting | .3 | .5 | .2 | | #+TBLFM: @2$>..@>>$> = ($2..$>> * trn(@>$2..@>$>>))_1

In words: set the last column of rows 2 through one before the last row equal to the first element of the vector (_1) that is the product of that row vector ($2..$>> i.e. cols 2 through one before the last column) with the transpose of the row vector in the last row and the same columns (@>$2..@>$>>).

The matrix product is a 1x1 matrix but calc interprets it as a vector: I first tried extracting the element using both row and column (_1_1) but the second _1 went unused, so I just deleted it.

Improve this answer
2
  • Surely the matrix product is 4 x 1, not 1 x 1? Is three some way to visualise the result of the calc expression before it gets substituted back into the org table? Commented Dec 2, 2019 at 23:14
  • It's not done all at once: it's done row by row, so it's the product of a 1x3 row vector (say the US one) and the transpose of another 1x3 row vector (the weights), so the result is indeed 1x1. You can turn on formula debugging (C-c {) and watch it step by step. Commented Dec 3, 2019 at 2:44

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.