Hi, Raul!
On 5/16/21, Raul Miller <[email protected]> wrote:
> You left out your example data for 'filer', but you gave enough detail
> that I think I can see what you are trying to do.
The following (which I did show) was the untransposed original
version, downloaded from Yahoo Finance:
Date Open High Low Close Volume
┌──────────┬──────┬──────┬──────┬──────┬────────┐
│2020-03-06│126.70│128.33│124.52│127.73│8193300 │
├──────────┼──────┼──────┼──────┼──────┼────────┤
│2020-03-09│120.16│122.41│117.28│117.81│10757500│
├──────────┼──────┼──────┼──────┼──────┼────────┤
│2020-03-10│122.78│124.88│115.76│124.77│11410700│
├──────────┼──────┼──────┼──────┼──────┼────────┤
│2020-03-11│121.01│122.58│116.38│117.97│8446500 │
├──────────┼──────┼──────┼──────┼──────┼────────┤
│2020-03-12│109.65│109.80│102.28│102.81│12512700│
└──────────┴──────┴──────┴──────┴──────┴────────┘
What I did then was to transpose it (and called it filer = file r) and
applied the following instructions which retained only the Date, High,
Low, and Close data:
filer=. 5 {. filer
filer=. 1 0 1 1 1 # filer
These were followed by the J code I showed in my previous email.
> That said, I think that I would split this up into two different routines:
>
> One routine would deal with unpacking the "rows" (or "inverted
> columns) from your data and amending the result. (And possibly either
> converting back to character or ensuring that both character and
> numeric arguments are supported or possibly not - for now, I'll follow
> the pattern from your example.)
>
> The other routine would deal with the numeric "rescaling". We could
> either have different routines here (one for each different kind of
> rescaling) or we could use a switch statement and pack them all
> together like you did. Either way works. This routine would be an
> argument for the other routine. (So the other routine would be an
> adverb.)
> ...
> hiloclose=: 0 0 1 1 1 0
>
> amender=:{{
> ndx=. I.x NB. index vector from bit vector
> update=. u 0&".@> x#y
> (<"0 update) ndx} y
> }}
>
> rescaler=:{{
> select. pricetype=. x
> case. 1 do. y
> case. 2 do. 10 * %: y
> case. 3 do. 100 * 10 ^. y
> end.
> }}
>
> Example use:
> hiloclose 1&rescaler amender filer
> hiloclose 2&rescaler amender filer
> hiloclose 3&rescaler amender filer
> ...
> Now... of course, you might actually want to avoid the amends that I
> assumed you wanted from your textual description. Or you might want
> other things to be different. So this is only an approximation of what
> you might do.
> ...
> And, of course, rather than have the three different rescaling verbs
> be 1&rescaler 2&rescaler and 3&rescaler you might instead give each of
> them a significant name.
>
> Still... perhaps this would be useful to you.
Thanks so much, Raul, for your response! I'm 75 and continue to learn
new things from replies to questions I've posed to this J programming
group. I have always loved learning during my entire life so far, and
J is a delightful challenge for my mental skills. Older folk like me
often work crossword puzzles or Sudoku to keep their minds sharp--I do
J programming for various computer software challenges! I programmed
as a hobby from 1975 to 2009, using various versions of BASIC over the
years. (I owned an original MITS Altair 8800, for which Bill Gates
and Paul Allen wrote a BASIC interpreter and thus started Microsoft.
Subsequently, I owned an Amiga 1000, and multiple IBM PC clones in
succeeding years.) In 2006 my library's IT department head suggested
that I might take a look at J, and, as they say, the rest is history.
Although I'm now generally comfortable programming in J (except for
occasional stumpers for which I need to ask for help in this group),
I'm at a point where I would like to learn how to write GUIs in J and
how to write visually oriented J programs (such as stock market
charting programs). I think it may be challenging because, from my
initial investigating, there seems to be *extremely* little
information (including coding samples) available about these aspects
of J programming.
Again, thanks so much, Raul!
Harvey
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
