Joe,
I am beginning to see why you are using boxopen so much: each cell is boxed
but must be unboxed for any calculations.
The trick of using r_webspread_ was really helpful.
I also see why calculations I wanted to test such as +/"1 r would not work.
Because r is the whole table, including the summation column (or row) that
the summation formula resides, the formula +/ &:> {. r in the rightmost
column will not work, but the formula +/ &:> {. 0 _1}.r does work.
So frustratingly, the cell at the intersection of a row and column that is
summarize/combine other cells, must be part of the array r, but the cell
must be excluded from the argument of the calculation.
I wonder if this suggest that instead of the output table being separate
from the input table, maybe a dummy calculation area which excludes the
current cell could be used temporarily?
On Sun, Feb 23, 2014 at 6:44 PM, Joe Bogner <[email protected]> wrote:
> On Sun, Feb 23, 2014 at 5:53 PM, Brian Schott <[email protected]
> >wrote:
>
> > Joe,
> >
> > I have webspread in a state where I always get the error message,
> > "ev_calc_click failed: Error: NETWORK_ERR: XMLHttpRequest Exception
> 101", I
> > think I fried webspread by entering (<1 2){r, or just r into the
> rightmost
> > cell of the first row because that 2 is bigger than the .
> >
> >
> Hey Brian, it's possible that something went awry with JHS when you were
> entering formulas. It's not a very robust, error-proof implementation of a
> spreadsheet. If in doubt, restart JHS
>
>
>
> > Before that I got no answer after typing +/}:0{:: r into any cell, but do
> > get an answer for +/>}:0{:: r . What is that > adding?
> >
> > It seems I need to add it to convert the unbox the array of numbers for
> the computation. I want each cell to either be a number or literal string
> (formula). See the comment below about inspecting the contents of the
> spreadsheet using r_webspread_
>
>
> > Also, the parens in r=:(".) :: (]) L:0 r seem to be unneeded.
> >
> >
> Thanks, I probably overuse them as I'm sometimes not sure when they are
> necessary
>
>
> > But I am getting frustrated with trying to make headway with
> > the limitations on the cell formulae.
> >
> >
> The best way to understand it is to use jijx in combination with the
> spreadsheet. I should have included that in the video. If you also have
> jijx (JHS) open, you can type r_webspread_ to see the r variable in the
> webspread locale. You can then test out various formulas in JHS easier
>
> Thanks for the feedback
>
> --
(B=)
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
