Very nice bob!

Just to add that as of J9, specifying an unboxed table as a selector is
supported so:
   rep (1 0 2 ,: 0 2 1)} t

┌──┬──┬──┬──┐

│0 │1 │2 │3 │

├──┼──┼──┼──┤

│4 │5 │6 │7 │

├──┼──┼──┼──┤

│8 │a │10│11│

└──┴──┴──┴──┘


┌──┬──┬──┬──┐

│12│13│a │15│

├──┼──┼──┼──┤

│16│17│18│19│

├──┼──┼──┼──┤

│20│21│22│23│

└──┴──┴──┴──┘

NB. and for a single replacement

rep (,: 1 0 2)} t

┌──┬──┬──┬──┐

│0 │1 │2 │3 │

├──┼──┼──┼──┤

│4 │5 │6 │7 │

├──┼──┼──┼──┤

│8 │9 │10│11│

└──┴──┴──┴──┘


┌──┬──┬──┬──┐

│12│13│a │15│

├──┼──┼──┼──┤

│16│17│18│19│

├──┼──┼──┼──┤

│20│21│22│23│

└──┴──┴──┴──┘

On Fri, Aug 7, 2020 at 4:20 PM 'robert therriault' via Programming <
[email protected]> wrote:

> Hi Harvey,
>
> Pascal got you most of the way there, but I will add the idea of using a
> selector to the Amend adverb to provide more precision.
>
>     [ t=: <"0 i. 2 3 4  NB. creates rank three boxed brick
> ┌──┬──┬──┬──┐
> │0 │1 │2 │3 │
> ├──┼──┼──┼──┤
> │4 │5 │6 │7 │
> ├──┼──┼──┼──┤
> │8 │9 │10│11│
> └──┴──┴──┴──┘
>
> ┌──┬──┬──┬──┐
> │12│13│14│15│
> ├──┼──┼──┼──┤
> │16│17│18│19│
> ├──┼──┼──┼──┤
> │20│21│22│23│
> └──┴──┴──┴──┘
>    [ rep=: <'a' NB. replacement value for selected addresses
> ┌─┐
> │a│
> └─┘
>    rep (1 }) t  NB. replaces all of the 1 item with rep's
> ┌─┬─┬──┬──┐
> │0│1│2 │3 │
> ├─┼─┼──┼──┤
> │4│5│6 │7 │
> ├─┼─┼──┼──┤
> │8│9│10│11│
> └─┴─┴──┴──┘
>
> ┌─┬─┬──┬──┐
> │a│a│a │a │
> ├─┼─┼──┼──┤
> │a│a│a │a │
> ├─┼─┼──┼──┤
> │a│a│a │a │
> └─┴─┴──┴──┘
>    rep ((<1 0) }) t  NB. replaces 0th row of the 1 item with rep's. (< 1
> 0) is necessary for selector rules
> ┌──┬──┬──┬──┐
> │0 │1 │2 │3 │
> ├──┼──┼──┼──┤
> │4 │5 │6 │7 │
> ├──┼──┼──┼──┤
> │8 │9 │10│11│
> └──┴──┴──┴──┘
>
> ┌──┬──┬──┬──┐
> │a │a │a │a │
> ├──┼──┼──┼──┤
> │16│17│18│19│
> ├──┼──┼──┼──┤
> │20│21│22│23│
> └──┴──┴──┴──┘
>    rep ((<1 0 2) }) t  NB. replaces  a single box in the 0th row of the 1
> item with rep. (< 1 0 2) is necessary for selector rules
> ┌──┬──┬──┬──┐
> │0 │1 │2 │3 │
> ├──┼──┼──┼──┤
> │4 │5 │6 │7 │
> ├──┼──┼──┼──┤
> │8 │9 │10│11│
> └──┴──┴──┴──┘
>
> ┌──┬──┬──┬──┐
> │12│13│a │15│
> ├──┼──┼──┼──┤
> │16│17│18│19│
> ├──┼──┼──┼──┤
> │20│21│22│23│
> └──┴──┴──┴──┘
>
> Hope this helps.
>
> Cheers, bob
>
> > On Aug 6, 2020, at 21:01, HH PackRat <[email protected]> wrote:
> >
> > On 8/6/20, 'Pascal Jasmin' via Programming <[email protected]>
> wrote:
> >>> NB.          (row{data)=. (< interpolated) start } row{data
> >>
> >> to change row 1,
> >> 2 2 2 2 2 (1}) i. 4 5
> >> 0 1 2 3 4
> >> 2 2 2 2 2
> >> 10 11 12 13 14
> >> 15 16 17 18 19
> >>
> >> if your data is boxed by column,
> >> 2 3 2 3 2  (1}) each <"1&.|: i. 4 5  NB. the new row can be a "boxed
> record"
> >
> > Thanks for your response, but your examples are the same kind of thing
> > I was encountering in the few sources that addressed amending.  It did
> > not address the issue of "drilling down" into the middle of a table
> > and changing a value.
> >
> > There are 3 levels of data: table (called "data"), row, and column.
> > Thus, every box of the table has a 3-part address.  Your example does
> > not indicate how to change a single cell in a table.  The diagram
> > below (sorry, I can only do a proportional font) shows the beginning
> > of the transposed data:
> >
> > ┌───────┬───────┬──────┬─┬──────┬───────┬─
> > │261.78      │263.33       │263.27    │0│263.23    │264.22      │
> > ├───────┼───────┼──────┼─┼──────┼───────┼─
> > │264.02      │264.59       │263.47    │0 │264.87   │265.17      │
> > ├───────┼───────┼──────┼─┼──────┼───────┼─
> > │261.54      │262.54      │262.84     │0 │262.62   │262.82      │
> > ├───────┼───────┼──────┼─┼──────┼───────┼─
> > │263.33      │263.27      │263.23     │0 │264.22    │263.92     │
> > ├───────┼───────┼──────┼─┼──────┼───────┼─
> > │1360000    │1150000    │300000   │0 │940000   │1040000    │
> > └───────┴───────┴──────┴─┴──────┴───────┴─
> >
> > How would you amend (in place) the entire data (table) so that the box
> > at column 3, row 2, would have the literal value '262.73' instead of
> > zero?  This is the essence of what I was trying to ask.
> >
> > Harvey
> > ----------------------------------------------------------------------
> > For information about J forums see http://www.jsoftware.com/forums.htm
>
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm
>
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm

Reply via email to